Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / db6a6a1c32ccc7345190d83d5668d4fa77cc5413 / . / docs / LangRef.html
blob: 35f83b48d4f54307fc465a75936999342be073de [file] [log] [blame]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
3<html>
4<head>
5 <title>LLVM Assembly Language Reference Manual</title>
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.">
10 <link rel="stylesheet" href="llvm.css" type="text/css">
11</head>
12
13<body>
14
15<div class="doc_title"> LLVM Language Reference Manual </div>
16<ol>
17 <li><a href="#abstract">Abstract</a></li>
18 <li><a href="#introduction">Introduction</a></li>
19 <li><a href="#identifiers">Identifiers</a></li>
20 <li><a href="#highlevel">High Level Structure</a>
21 <ol>
22 <li><a href="#modulestructure">Module Structure</a></li>
23 <li><a href="#linkage">Linkage Types</a></li>
24 <li><a href="#callingconv">Calling Conventions</a></li>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]25 <li><a href="#namedtypes">Named Types</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]26 <li><a href="#globalvars">Global Variables</a></li>
27 <li><a href="#functionstructure">Functions</a></li>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]28 <li><a href="#aliasstructure">Aliases</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]29 <li><a href="#paramattrs">Parameter Attributes</a></li>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]30 <li><a href="#fnattrs">Function Attributes</a></li>
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]31 <li><a href="#gc">Garbage Collector Names</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]32 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
33 <li><a href="#datalayout">Data Layout</a></li>
34 </ol>
35 </li>
36 <li><a href="#typesystem">Type System</a>
37 <ol>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]38 <li><a href="#t_classifications">Type Classifications</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]39 <li><a href="#t_primitive">Primitive Types</a>
40 <ol>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]41 <li><a href="#t_floating">Floating Point Types</a></li>
42 <li><a href="#t_void">Void Type</a></li>
43 <li><a href="#t_label">Label Type</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]44 </ol>
45 </li>
46 <li><a href="#t_derived">Derived Types</a>
47 <ol>
Chris Lattner251ab812007-12-18 06:18:21 +0000[diff] [blame]48 <li><a href="#t_integer">Integer Type</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]49 <li><a href="#t_array">Array Type</a></li>
50 <li><a href="#t_function">Function Type</a></li>
51 <li><a href="#t_pointer">Pointer Type</a></li>
52 <li><a href="#t_struct">Structure Type</a></li>
53 <li><a href="#t_pstruct">Packed Structure Type</a></li>
54 <li><a href="#t_vector">Vector Type</a></li>
55 <li><a href="#t_opaque">Opaque Type</a></li>
56 </ol>
57 </li>
58 </ol>
59 </li>
60 <li><a href="#constants">Constants</a>
61 <ol>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]62 <li><a href="#simpleconstants">Simple Constants</a></li>
63 <li><a href="#aggregateconstants">Aggregate Constants</a></li>
64 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
65 <li><a href="#undefvalues">Undefined Values</a></li>
66 <li><a href="#constantexprs">Constant Expressions</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]67 </ol>
68 </li>
69 <li><a href="#othervalues">Other Values</a>
70 <ol>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]71 <li><a href="#inlineasm">Inline Assembler Expressions</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]72 </ol>
73 </li>
74 <li><a href="#instref">Instruction Reference</a>
75 <ol>
76 <li><a href="#terminators">Terminator Instructions</a>
77 <ol>
78 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
79 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
80 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
81 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
82 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
83 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
84 </ol>
85 </li>
86 <li><a href="#binaryops">Binary Operations</a>
87 <ol>
88 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
89 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
90 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
91 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
92 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
93 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
94 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
95 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
96 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
97 </ol>
98 </li>
99 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
100 <ol>
101 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
102 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
103 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
104 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
105 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
106 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
107 </ol>
108 </li>
109 <li><a href="#vectorops">Vector Operations</a>
110 <ol>
111 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
112 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
113 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
114 </ol>
115 </li>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]116 <li><a href="#aggregateops">Aggregate Operations</a>
117 <ol>
118 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
119 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
120 </ol>
121 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]122 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
123 <ol>
124 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
125 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
126 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
127 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
128 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
129 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
130 </ol>
131 </li>
132 <li><a href="#convertops">Conversion Operations</a>
133 <ol>
134 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
135 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
136 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
137 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
138 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
139 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
140 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
141 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
142 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
143 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
144 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
145 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
146 </ol>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]147 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]148 <li><a href="#otherops">Other Operations</a>
149 <ol>
150 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
151 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]152 <li><a href="#i_vicmp">'<tt>vicmp</tt>' Instruction</a></li>
153 <li><a href="#i_vfcmp">'<tt>vfcmp</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]154 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
155 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
156 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
157 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
158 </ol>
159 </li>
160 </ol>
161 </li>
162 <li><a href="#intrinsics">Intrinsic Functions</a>
163 <ol>
164 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
165 <ol>
166 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
167 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
168 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
169 </ol>
170 </li>
171 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
172 <ol>
173 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
174 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
175 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
176 </ol>
177 </li>
178 <li><a href="#int_codegen">Code Generator Intrinsics</a>
179 <ol>
180 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
181 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
182 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
183 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
184 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
185 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
186 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
187 </ol>
188 </li>
189 <li><a href="#int_libc">Standard C Library Intrinsics</a>
190 <ol>
191 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
192 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
193 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
194 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
195 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]196 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
197 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
198 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]199 </ol>
200 </li>
201 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
202 <ol>
203 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
204 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
205 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
206 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
207 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
208 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
209 </ol>
210 </li>
211 <li><a href="#int_debugger">Debugger intrinsics</a></li>
212 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]213 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]214 <ol>
215 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]216 </ol>
217 </li>
Bill Wendling9127adb2008-11-18 22:10:53 +0000[diff] [blame]218 <li><a href="#int_atomics">Atomic intrinsics</a>
219 <ol>
220 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
221 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
222 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
223 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
224 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
225 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
226 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
227 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
228 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
229 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
230 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
231 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
232 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
233 </ol>
234 </li>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]235 <li><a href="#int_general">General intrinsics</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]236 <ol>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]237 <li><a href="#int_var_annotation">
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]238 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]239 <li><a href="#int_annotation">
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]240 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]241 <li><a href="#int_trap">
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]242 '<tt>llvm.trap</tt>' Intrinsic</a></li>
243 <li><a href="#int_stackprotector">
244 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]245 </ol>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]246 </li>
247 </ol>
248 </li>
249</ol>
250
251<div class="doc_author">
252 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
253 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
254</div>
255
256<!-- *********************************************************************** -->
257<div class="doc_section"> <a name="abstract">Abstract </a></div>
258<!-- *********************************************************************** -->
259
260<div class="doc_text">
261<p>This document is a reference manual for the LLVM assembly language.
Bill Wendlinge7846a52008-08-05 22:29:16 +0000[diff] [blame]262LLVM is a Static Single Assignment (SSA) based representation that provides
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]263type safety, low-level operations, flexibility, and the capability of
264representing 'all' high-level languages cleanly. It is the common code
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]265representation used throughout all phases of the LLVM compilation
266strategy.</p>
267</div>
268
269<!-- *********************************************************************** -->
270<div class="doc_section"> <a name="introduction">Introduction</a> </div>
271<!-- *********************************************************************** -->
272
273<div class="doc_text">
274
275<p>The LLVM code representation is designed to be used in three
276different forms: as an in-memory compiler IR, as an on-disk bitcode
277representation (suitable for fast loading by a Just-In-Time compiler),
278and as a human readable assembly language representation. This allows
279LLVM to provide a powerful intermediate representation for efficient
280compiler transformations and analysis, while providing a natural means
281to debug and visualize the transformations. The three different forms
282of LLVM are all equivalent. This document describes the human readable
283representation and notation.</p>
284
285<p>The LLVM representation aims to be light-weight and low-level
286while being expressive, typed, and extensible at the same time. It
287aims to be a "universal IR" of sorts, by being at a low enough level
288that high-level ideas may be cleanly mapped to it (similar to how
289microprocessors are "universal IR's", allowing many source languages to
290be mapped to them). By providing type information, LLVM can be used as
291the target of optimizations: for example, through pointer analysis, it
292can be proven that a C automatic variable is never accessed outside of
293the current function... allowing it to be promoted to a simple SSA
294value instead of a memory location.</p>
295
296</div>
297
298<!-- _______________________________________________________________________ -->
299<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
300
301<div class="doc_text">
302
303<p>It is important to note that this document describes 'well formed'
304LLVM assembly language. There is a difference between what the parser
305accepts and what is considered 'well formed'. For example, the
306following instruction is syntactically okay, but not well formed:</p>
307
308<div class="doc_code">
309<pre>
310%x = <a href="#i_add">add</a> i32 1, %x
311</pre>
312</div>
313
314<p>...because the definition of <tt>%x</tt> does not dominate all of
315its uses. The LLVM infrastructure provides a verification pass that may
316be used to verify that an LLVM module is well formed. This pass is
317automatically run by the parser after parsing input assembly and by
318the optimizer before it outputs bitcode. The violations pointed out
319by the verifier pass indicate bugs in transformation passes or input to
320the parser.</p>
321</div>
322
Chris Lattnera83fdc02007-10-03 17:34:29 +0000[diff] [blame]323<!-- Describe the typesetting conventions here. -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]324
325<!-- *********************************************************************** -->
326<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
327<!-- *********************************************************************** -->
328
329<div class="doc_text">
330
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]331 <p>LLVM identifiers come in two basic types: global and local. Global
332 identifiers (functions, global variables) begin with the @ character. Local
333 identifiers (register names, types) begin with the % character. Additionally,
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]334 there are three different formats for identifiers, for different purposes:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]335
336<ol>
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]337 <li>Named values are represented as a string of characters with their prefix.
338 For example, %foo, @DivisionByZero, %a.really.long.identifier. The actual
339 regular expression used is '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]340 Identifiers which require other characters in their names can be surrounded
Daniel Dunbar73f9a772008-10-14 23:51:43 +0000[diff] [blame]341 with quotes. Special characters may be escaped using "\xx" where xx is the
342 ASCII code for the character in hexadecimal. In this way, any character can
343 be used in a name value, even quotes themselves.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]344
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]345 <li>Unnamed values are represented as an unsigned numeric value with their
346 prefix. For example, %12, @2, %44.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]347
348 <li>Constants, which are described in a <a href="#constants">section about
349 constants</a>, below.</li>
350</ol>
351
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]352<p>LLVM requires that values start with a prefix for two reasons: Compilers
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]353don't need to worry about name clashes with reserved words, and the set of
354reserved words may be expanded in the future without penalty. Additionally,
355unnamed identifiers allow a compiler to quickly come up with a temporary
356variable without having to avoid symbol table conflicts.</p>
357
358<p>Reserved words in LLVM are very similar to reserved words in other
359languages. There are keywords for different opcodes
360('<tt><a href="#i_add">add</a></tt>',
361 '<tt><a href="#i_bitcast">bitcast</a></tt>',
362 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
363href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
364and others. These reserved words cannot conflict with variable names, because
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]365none of them start with a prefix character ('%' or '@').</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]366
367<p>Here is an example of LLVM code to multiply the integer variable
368'<tt>%X</tt>' by 8:</p>
369
370<p>The easy way:</p>
371
372<div class="doc_code">
373<pre>
374%result = <a href="#i_mul">mul</a> i32 %X, 8
375</pre>
376</div>
377
378<p>After strength reduction:</p>
379
380<div class="doc_code">
381<pre>
382%result = <a href="#i_shl">shl</a> i32 %X, i8 3
383</pre>
384</div>
385
386<p>And the hard way:</p>
387
388<div class="doc_code">
389<pre>
390<a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
391<a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
392%result = <a href="#i_add">add</a> i32 %1, %1
393</pre>
394</div>
395
396<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
397important lexical features of LLVM:</p>
398
399<ol>
400
401 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
402 line.</li>
403
404 <li>Unnamed temporaries are created when the result of a computation is not
405 assigned to a named value.</li>
406
407 <li>Unnamed temporaries are numbered sequentially</li>
408
409</ol>
410
411<p>...and it also shows a convention that we follow in this document. When
412demonstrating instructions, we will follow an instruction with a comment that
413defines the type and name of value produced. Comments are shown in italic
414text.</p>
415
416</div>
417
418<!-- *********************************************************************** -->
419<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
420<!-- *********************************************************************** -->
421
422<!-- ======================================================================= -->
423<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
424</div>
425
426<div class="doc_text">
427
428<p>LLVM programs are composed of "Module"s, each of which is a
429translation unit of the input programs. Each module consists of
430functions, global variables, and symbol table entries. Modules may be
431combined together with the LLVM linker, which merges function (and
432global variable) definitions, resolves forward declarations, and merges
433symbol table entries. Here is an example of the "hello world" module:</p>
434
435<div class="doc_code">
436<pre><i>; Declare the string constant as a global constant...</i>
437<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a
438 href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>
439
440<i>; External declaration of the puts function</i>
441<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
442
443<i>; Definition of main function</i>
444define i32 @main() { <i>; i32()* </i>
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]445 <i>; Convert [13 x i8]* to i8 *...</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]446 %cast210 = <a
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]447 href="#i_getelementptr">getelementptr</a> [13 x i8]* @.LC0, i64 0, i64 0 <i>; i8 *</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]448
449 <i>; Call puts function to write out the string to stdout...</i>
450 <a
451 href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
452 <a
453 href="#i_ret">ret</a> i32 0<br>}<br>
454</pre>
455</div>
456
457<p>This example is made up of a <a href="#globalvars">global variable</a>
458named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
459function, and a <a href="#functionstructure">function definition</a>
460for "<tt>main</tt>".</p>
461
462<p>In general, a module is made up of a list of global values,
463where both functions and global variables are global values. Global values are
464represented by a pointer to a memory location (in this case, a pointer to an
465array of char, and a pointer to a function), and have one of the following <a
466href="#linkage">linkage types</a>.</p>
467
468</div>
469
470<!-- ======================================================================= -->
471<div class="doc_subsection">
472 <a name="linkage">Linkage Types</a>
473</div>
474
475<div class="doc_text">
476
477<p>
478All Global Variables and Functions have one of the following types of linkage:
479</p>
480
481<dl>
482
Rafael Espindolaa168fc92009-01-15 20:18:42 +0000[diff] [blame]483 <dt><tt><b><a name="linkage_private">private</a></b></tt>: </dt>
484
485 <dd>Global values with private linkage are only directly accessible by
486 objects in the current module. In particular, linking code into a module with
487 an private global value may cause the private to be renamed as necessary to
488 avoid collisions. Because the symbol is private to the module, all
489 references can be updated. This doesn't show up in any symbol table in the
490 object file.
491 </dd>
492
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame]493 <dt><tt><b><a name="linkage_internal">internal</a></b></tt>: </dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]494
Duncan Sandsa75223a2009-01-16 09:29:46 +0000[diff] [blame]495 <dd> Similar to private, but the value shows as a local symbol (STB_LOCAL in
Rafael Espindolaa168fc92009-01-15 20:18:42 +0000[diff] [blame]496 the case of ELF) in the object file. This corresponds to the notion of the
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]497 '<tt>static</tt>' keyword in C.
498 </dd>
499
500 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
501
502 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
503 the same name when linkage occurs. This is typically used to implement
504 inline functions, templates, or other code which must be generated in each
505 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
506 allowed to be discarded.
507 </dd>
508
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame]509 <dt><tt><b><a name="linkage_common">common</a></b></tt>: </dt>
510
511 <dd>"<tt>common</tt>" linkage is exactly the same as <tt>linkonce</tt>
512 linkage, except that unreferenced <tt>common</tt> globals may not be
513 discarded. This is used for globals that may be emitted in multiple
514 translation units, but that are not guaranteed to be emitted into every
515 translation unit that uses them. One example of this is tentative
516 definitions in C, such as "<tt>int X;</tt>" at global scope.
517 </dd>
518
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]519 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
520
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame]521 <dd>"<tt>weak</tt>" linkage is the same as <tt>common</tt> linkage, except
522 that some targets may choose to emit different assembly sequences for them
523 for target-dependent reasons. This is used for globals that are declared
524 "weak" in C source code.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]525 </dd>
526
527 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
528
529 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
530 pointer to array type. When two global variables with appending linkage are
531 linked together, the two global arrays are appended together. This is the
532 LLVM, typesafe, equivalent of having the system linker append together
533 "sections" with identical names when .o files are linked.
534 </dd>
535
536 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]537 <dd>The semantics of this linkage follow the ELF object file model: the
538 symbol is weak until linked, if not linked, the symbol becomes null instead
539 of being an undefined reference.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]540 </dd>
541
542 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
543
544 <dd>If none of the above identifiers are used, the global is externally
545 visible, meaning that it participates in linkage and can be used to resolve
546 external symbol references.
547 </dd>
548</dl>
549
550 <p>
551 The next two types of linkage are targeted for Microsoft Windows platform
552 only. They are designed to support importing (exporting) symbols from (to)
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]553 DLLs (Dynamic Link Libraries).
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]554 </p>
555
556 <dl>
557 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
558
559 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
560 or variable via a global pointer to a pointer that is set up by the DLL
561 exporting the symbol. On Microsoft Windows targets, the pointer name is
Dan Gohman5ec99832009-01-12 21:35:55 +0000[diff] [blame]562 formed by combining <code>__imp_</code> and the function or variable name.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]563 </dd>
564
565 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
566
567 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
568 pointer to a pointer in a DLL, so that it can be referenced with the
569 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
Dan Gohman5ec99832009-01-12 21:35:55 +0000[diff] [blame]570 name is formed by combining <code>__imp_</code> and the function or variable
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]571 name.
572 </dd>
573
574</dl>
575
Dan Gohman4dfac702008-11-24 17:18:39 +0000[diff] [blame]576<p>For example, since the "<tt>.LC0</tt>"
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]577variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
578variable and was linked with this one, one of the two would be renamed,
579preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
580external (i.e., lacking any linkage declarations), they are accessible
581outside of the current module.</p>
582<p>It is illegal for a function <i>declaration</i>
583to have any linkage type other than "externally visible", <tt>dllimport</tt>,
584or <tt>extern_weak</tt>.</p>
585<p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]586linkages.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]587</div>
588
589<!-- ======================================================================= -->
590<div class="doc_subsection">
591 <a name="callingconv">Calling Conventions</a>
592</div>
593
594<div class="doc_text">
595
596<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
597and <a href="#i_invoke">invokes</a> can all have an optional calling convention
598specified for the call. The calling convention of any pair of dynamic
599caller/callee must match, or the behavior of the program is undefined. The
600following calling conventions are supported by LLVM, and more may be added in
601the future:</p>
602
603<dl>
604 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
605
606 <dd>This calling convention (the default if no other calling convention is
607 specified) matches the target C calling conventions. This calling convention
608 supports varargs function calls and tolerates some mismatch in the declared
609 prototype and implemented declaration of the function (as does normal C).
610 </dd>
611
612 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
613
614 <dd>This calling convention attempts to make calls as fast as possible
615 (e.g. by passing things in registers). This calling convention allows the
616 target to use whatever tricks it wants to produce fast code for the target,
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]617 without having to conform to an externally specified ABI (Application Binary
618 Interface). Implementations of this convention should allow arbitrary
Arnold Schwaighofer07444922008-05-14 09:17:12 +0000[diff] [blame]619 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> to be
620 supported. This calling convention does not support varargs and requires the
621 prototype of all callees to exactly match the prototype of the function
622 definition.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]623 </dd>
624
625 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
626
627 <dd>This calling convention attempts to make code in the caller as efficient
628 as possible under the assumption that the call is not commonly executed. As
629 such, these calls often preserve all registers so that the call does not break
630 any live ranges in the caller side. This calling convention does not support
631 varargs and requires the prototype of all callees to exactly match the
632 prototype of the function definition.
633 </dd>
634
635 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
636
637 <dd>Any calling convention may be specified by number, allowing
638 target-specific calling conventions to be used. Target specific calling
639 conventions start at 64.
640 </dd>
641</dl>
642
643<p>More calling conventions can be added/defined on an as-needed basis, to
644support pascal conventions or any other well-known target-independent
645convention.</p>
646
647</div>
648
649<!-- ======================================================================= -->
650<div class="doc_subsection">
651 <a name="visibility">Visibility Styles</a>
652</div>
653
654<div class="doc_text">
655
656<p>
657All Global Variables and Functions have one of the following visibility styles:
658</p>
659
660<dl>
661 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
662
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]663 <dd>On targets that use the ELF object file format, default visibility means
664 that the declaration is visible to other
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]665 modules and, in shared libraries, means that the declared entity may be
666 overridden. On Darwin, default visibility means that the declaration is
667 visible to other modules. Default visibility corresponds to "external
668 linkage" in the language.
669 </dd>
670
671 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
672
673 <dd>Two declarations of an object with hidden visibility refer to the same
674 object if they are in the same shared object. Usually, hidden visibility
675 indicates that the symbol will not be placed into the dynamic symbol table,
676 so no other module (executable or shared library) can reference it
677 directly.
678 </dd>
679
680 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
681
682 <dd>On ELF, protected visibility indicates that the symbol will be placed in
683 the dynamic symbol table, but that references within the defining module will
684 bind to the local symbol. That is, the symbol cannot be overridden by another
685 module.
686 </dd>
687</dl>
688
689</div>
690
691<!-- ======================================================================= -->
692<div class="doc_subsection">
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]693 <a name="namedtypes">Named Types</a>
694</div>
695
696<div class="doc_text">
697
698<p>LLVM IR allows you to specify name aliases for certain types. This can make
699it easier to read the IR and make the IR more condensed (particularly when
700recursive types are involved). An example of a name specification is:
701</p>
702
703<div class="doc_code">
704<pre>
705%mytype = type { %mytype*, i32 }
706</pre>
707</div>
708
709<p>You may give a name to any <a href="#typesystem">type</a> except "<a
710href="t_void">void</a>". Type name aliases may be used anywhere a type is
711expected with the syntax "%mytype".</p>
712
713<p>Note that type names are aliases for the structural type that they indicate,
714and that you can therefore specify multiple names for the same type. This often
715leads to confusing behavior when dumping out a .ll file. Since LLVM IR uses
716structural typing, the name is not part of the type. When printing out LLVM IR,
717the printer will pick <em>one name</em> to render all types of a particular
718shape. This means that if you have code where two different source types end up
719having the same LLVM type, that the dumper will sometimes print the "wrong" or
720unexpected type. This is an important design point and isn't going to
721change.</p>
722
723</div>
724
725
726<!-- ======================================================================= -->
727<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]728 <a name="globalvars">Global Variables</a>
729</div>
730
731<div class="doc_text">
732
733<p>Global variables define regions of memory allocated at compilation time
734instead of run-time. Global variables may optionally be initialized, may have
735an explicit section to be placed in, and may have an optional explicit alignment
736specified. A variable may be defined as "thread_local", which means that it
737will not be shared by threads (each thread will have a separated copy of the
738variable). A variable may be defined as a global "constant," which indicates
739that the contents of the variable will <b>never</b> be modified (enabling better
740optimization, allowing the global data to be placed in the read-only section of
741an executable, etc). Note that variables that need runtime initialization
742cannot be marked "constant" as there is a store to the variable.</p>
743
744<p>
745LLVM explicitly allows <em>declarations</em> of global variables to be marked
746constant, even if the final definition of the global is not. This capability
747can be used to enable slightly better optimization of the program, but requires
748the language definition to guarantee that optimizations based on the
749'constantness' are valid for the translation units that do not include the
750definition.
751</p>
752
753<p>As SSA values, global variables define pointer values that are in
754scope (i.e. they dominate) all basic blocks in the program. Global
755variables always define a pointer to their "content" type because they
756describe a region of memory, and all memory objects in LLVM are
757accessed through pointers.</p>
758
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]759<p>A global variable may be declared to reside in a target-specifc numbered
760address space. For targets that support them, address spaces may affect how
761optimizations are performed and/or what target instructions are used to access
Christopher Lamb20a39e92007-12-12 08:44:39 +0000[diff] [blame]762the variable. The default address space is zero. The address space qualifier
763must precede any other attributes.</p>
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]764
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]765<p>LLVM allows an explicit section to be specified for globals. If the target
766supports it, it will emit globals to the section specified.</p>
767
768<p>An explicit alignment may be specified for a global. If not present, or if
769the alignment is set to zero, the alignment of the global is set by the target
770to whatever it feels convenient. If an explicit alignment is specified, the
771global is forced to have at least that much alignment. All alignments must be
772a power of 2.</p>
773
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]774<p>For example, the following defines a global in a numbered address space with
775an initializer, section, and alignment:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]776
777<div class="doc_code">
778<pre>
Dan Gohman21ef02c2009-01-11 00:40:00 +0000[diff] [blame]779@G = addrspace(5) constant float 1.0, section "foo", align 4
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]780</pre>
781</div>
782
783</div>
784
785
786<!-- ======================================================================= -->
787<div class="doc_subsection">
788 <a name="functionstructure">Functions</a>
789</div>
790
791<div class="doc_text">
792
793<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
794an optional <a href="#linkage">linkage type</a>, an optional
795<a href="#visibility">visibility style</a>, an optional
796<a href="#callingconv">calling convention</a>, a return type, an optional
797<a href="#paramattrs">parameter attribute</a> for the return type, a function
798name, a (possibly empty) argument list (each with optional
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]799<a href="#paramattrs">parameter attributes</a>), optional
800<a href="#fnattrs">function attributes</a>, an optional section,
801an optional alignment, an optional <a href="#gc">garbage collector name</a>,
Chris Lattneree202542008-10-04 18:10:21 +0000[diff] [blame]802an opening curly brace, a list of basic blocks, and a closing curly brace.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]803
804LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
805optional <a href="#linkage">linkage type</a>, an optional
806<a href="#visibility">visibility style</a>, an optional
807<a href="#callingconv">calling convention</a>, a return type, an optional
808<a href="#paramattrs">parameter attribute</a> for the return type, a function
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]809name, a possibly empty list of arguments, an optional alignment, and an optional
Gordon Henriksend606f5b2007-12-10 03:30:21 +0000[diff] [blame]810<a href="#gc">garbage collector name</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]811
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]812<p>A function definition contains a list of basic blocks, forming the CFG
813(Control Flow Graph) for
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]814the function. Each basic block may optionally start with a label (giving the
815basic block a symbol table entry), contains a list of instructions, and ends
816with a <a href="#terminators">terminator</a> instruction (such as a branch or
817function return).</p>
818
819<p>The first basic block in a function is special in two ways: it is immediately
820executed on entrance to the function, and it is not allowed to have predecessor
821basic blocks (i.e. there can not be any branches to the entry block of a
822function). Because the block can have no predecessors, it also cannot have any
823<a href="#i_phi">PHI nodes</a>.</p>
824
825<p>LLVM allows an explicit section to be specified for functions. If the target
826supports it, it will emit functions to the section specified.</p>
827
828<p>An explicit alignment may be specified for a function. If not present, or if
829the alignment is set to zero, the alignment of the function is set by the target
830to whatever it feels convenient. If an explicit alignment is specified, the
831function is forced to have at least that much alignment. All alignments must be
832a power of 2.</p>
833
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]834 <h5>Syntax:</h5>
835
836<div class="doc_code">
Chris Lattner1e5c5cd02008-10-13 16:55:18 +0000[diff] [blame]837<tt>
838define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
839 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
840 &lt;ResultType&gt; @&lt;FunctionName&gt; ([argument list])
841 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
842 [<a href="#gc">gc</a>] { ... }
843</tt>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]844</div>
845
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]846</div>
847
848
849<!-- ======================================================================= -->
850<div class="doc_subsection">
851 <a name="aliasstructure">Aliases</a>
852</div>
853<div class="doc_text">
854 <p>Aliases act as "second name" for the aliasee value (which can be either
Anton Korobeynikov96822812008-03-22 08:36:14 +0000[diff] [blame]855 function, global variable, another alias or bitcast of global value). Aliases
856 may have an optional <a href="#linkage">linkage type</a>, and an
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]857 optional <a href="#visibility">visibility style</a>.</p>
858
859 <h5>Syntax:</h5>
860
861<div class="doc_code">
862<pre>
Duncan Sandsd7bfabf2008-09-12 20:48:21 +0000[diff] [blame]863@&lt;Name&gt; = alias [Linkage] [Visibility] &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]864</pre>
865</div>
866
867</div>
868
869
870
871<!-- ======================================================================= -->
872<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
873<div class="doc_text">
874 <p>The return type and each parameter of a function type may have a set of
875 <i>parameter attributes</i> associated with them. Parameter attributes are
876 used to communicate additional information about the result or parameters of
Duncan Sandsf5588dc2007-11-27 13:23:08 +0000[diff] [blame]877 a function. Parameter attributes are considered to be part of the function,
878 not of the function type, so functions with different parameter attributes
879 can have the same function type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]880
881 <p>Parameter attributes are simple keywords that follow the type specified. If
882 multiple parameter attributes are needed, they are space separated. For
883 example:</p>
884
885<div class="doc_code">
886<pre>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]887declare i32 @printf(i8* noalias , ...)
Chris Lattnerf33b8452008-10-04 18:33:34 +0000[diff] [blame]888declare i32 @atoi(i8 zeroext)
889declare signext i8 @returns_signed_char()
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]890</pre>
891</div>
892
Duncan Sandsf5588dc2007-11-27 13:23:08 +0000[diff] [blame]893 <p>Note that any attributes for the function result (<tt>nounwind</tt>,
894 <tt>readonly</tt>) come immediately after the argument list.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]895
896 <p>Currently, only the following parameter attributes are defined:</p>
897 <dl>
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]898 <dt><tt>zeroext</tt></dt>
Chris Lattnerf33b8452008-10-04 18:33:34 +0000[diff] [blame]899 <dd>This indicates to the code generator that the parameter or return value
900 should be zero-extended to a 32-bit value by the caller (for a parameter)
901 or the callee (for a return value).</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]902
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]903 <dt><tt>signext</tt></dt>
Chris Lattnerf33b8452008-10-04 18:33:34 +0000[diff] [blame]904 <dd>This indicates to the code generator that the parameter or return value
905 should be sign-extended to a 32-bit value by the caller (for a parameter)
906 or the callee (for a return value).</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]907
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]908 <dt><tt>inreg</tt></dt>
Dale Johannesenc08a0e22008-09-25 20:47:45 +0000[diff] [blame]909 <dd>This indicates that this parameter or return value should be treated
910 in a special target-dependent fashion during while emitting code for a
911 function call or return (usually, by putting it in a register as opposed
Chris Lattnerf33b8452008-10-04 18:33:34 +0000[diff] [blame]912 to memory, though some targets use it to distinguish between two different
913 kinds of registers). Use of this attribute is target-specific.</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]914
Duncan Sands7d94e5a2008-10-06 08:14:18 +0000[diff] [blame]915 <dt><tt><a name="byval">byval</a></tt></dt>
Chris Lattner04c86182008-01-15 04:34:22 +0000[diff] [blame]916 <dd>This indicates that the pointer parameter should really be passed by
917 value to the function. The attribute implies that a hidden copy of the
918 pointee is made between the caller and the callee, so the callee is unable
Chris Lattner6a9f3c42008-08-05 18:21:08 +0000[diff] [blame]919 to modify the value in the callee. This attribute is only valid on LLVM
Chris Lattner04c86182008-01-15 04:34:22 +0000[diff] [blame]920 pointer arguments. It is generally used to pass structs and arrays by
Duncan Sands7d94e5a2008-10-06 08:14:18 +0000[diff] [blame]921 value, but is also valid on pointers to scalars. The copy is considered to
922 belong to the caller not the callee (for example,
923 <tt><a href="#readonly">readonly</a></tt> functions should not write to
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]924 <tt>byval</tt> parameters). This is not a valid attribute for return
925 values. </dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]926
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]927 <dt><tt>sret</tt></dt>
Duncan Sands616cc032008-02-18 04:19:38 +0000[diff] [blame]928 <dd>This indicates that the pointer parameter specifies the address of a
929 structure that is the return value of the function in the source program.
Chris Lattnerf33b8452008-10-04 18:33:34 +0000[diff] [blame]930 This pointer must be guaranteed by the caller to be valid: loads and stores
931 to the structure may be assumed by the callee to not to trap. This may only
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]932 be applied to the first parameter. This is not a valid attribute for
933 return values. </dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]934
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]935 <dt><tt>noalias</tt></dt>
Nick Lewyckyff384472008-11-24 03:41:24 +0000[diff] [blame]936 <dd>This indicates that the pointer does not alias any global or any other
937 parameter. The caller is responsible for ensuring that this is the
Nick Lewyckya604a942008-11-24 05:00:44 +0000[diff] [blame]938 case. On a function return value, <tt>noalias</tt> additionally indicates
939 that the pointer does not alias any other pointers visible to the
Nick Lewycky40fb6f22008-12-19 06:39:12 +0000[diff] [blame]940 caller. For further details, please see the discussion of the NoAlias
941 response in
942 <a href="http://llvm.org/docs/AliasAnalysis.html#MustMayNo">alias
943 analysis</a>.</dd>
944
945 <dt><tt>nocapture</tt></dt>
946 <dd>This indicates that the callee does not make any copies of the pointer
947 that outlive the callee itself. This is not a valid attribute for return
948 values.</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]949
Duncan Sands4ee46812007-07-27 19:57:41 +0000[diff] [blame]950 <dt><tt>nest</tt></dt>
Duncan Sandsf1a7d4c2008-07-08 09:27:25 +0000[diff] [blame]951 <dd>This indicates that the pointer parameter can be excised using the
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]952 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
953 attribute for return values.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]954 </dl>
955
956</div>
957
958<!-- ======================================================================= -->
959<div class="doc_subsection">
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]960 <a name="gc">Garbage Collector Names</a>
961</div>
962
963<div class="doc_text">
964<p>Each function may specify a garbage collector name, which is simply a
965string.</p>
966
967<div class="doc_code"><pre
968>define void @f() gc "name" { ...</pre></div>
969
970<p>The compiler declares the supported values of <i>name</i>. Specifying a
971collector which will cause the compiler to alter its output in order to support
972the named garbage collection algorithm.</p>
973</div>
974
975<!-- ======================================================================= -->
976<div class="doc_subsection">
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]977 <a name="fnattrs">Function Attributes</a>
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]978</div>
979
980<div class="doc_text">
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]981
982<p>Function attributes are set to communicate additional information about
983 a function. Function attributes are considered to be part of the function,
984 not of the function type, so functions with different parameter attributes
985 can have the same function type.</p>
986
987 <p>Function attributes are simple keywords that follow the type specified. If
988 multiple attributes are needed, they are space separated. For
989 example:</p>
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]990
991<div class="doc_code">
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]992<pre>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]993define void @f() noinline { ... }
994define void @f() alwaysinline { ... }
995define void @f() alwaysinline optsize { ... }
996define void @f() optsize
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]997</pre>
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]998</div>
999
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1000<dl>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1001<dt><tt>alwaysinline</tt></dt>
Chris Lattner82d70f92008-10-04 18:23:17 +0000[diff] [blame]1002<dd>This attribute indicates that the inliner should attempt to inline this
1003function into callers whenever possible, ignoring any active inlining size
1004threshold for this caller.</dd>
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1005
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1006<dt><tt>noinline</tt></dt>
Chris Lattner82d70f92008-10-04 18:23:17 +0000[diff] [blame]1007<dd>This attribute indicates that the inliner should never inline this function
Chris Lattner377f3ae2008-10-05 17:14:59 +0000[diff] [blame]1008in any situation. This attribute may not be used together with the
Chris Lattner82d70f92008-10-04 18:23:17 +0000[diff] [blame]1009<tt>alwaysinline</tt> attribute.</dd>
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1010
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1011<dt><tt>optsize</tt></dt>
Devang Patelc29099b2008-09-29 18:34:44 +0000[diff] [blame]1012<dd>This attribute suggests that optimization passes and code generator passes
Chris Lattner82d70f92008-10-04 18:23:17 +0000[diff] [blame]1013make choices that keep the code size of this function low, and otherwise do
1014optimizations specifically to reduce code size.</dd>
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1015
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1016<dt><tt>noreturn</tt></dt>
Chris Lattner82d70f92008-10-04 18:23:17 +0000[diff] [blame]1017<dd>This function attribute indicates that the function never returns normally.
1018This produces undefined behavior at runtime if the function ever does
1019dynamically return.</dd>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1020
1021<dt><tt>nounwind</tt></dt>
Chris Lattner82d70f92008-10-04 18:23:17 +0000[diff] [blame]1022<dd>This function attribute indicates that the function never returns with an
1023unwind or exceptional control flow. If the function does unwind, its runtime
1024behavior is undefined.</dd>
1025
1026<dt><tt>readnone</tt></dt>
Duncan Sands7d94e5a2008-10-06 08:14:18 +0000[diff] [blame]1027<dd>This attribute indicates that the function computes its result (or the
1028exception it throws) based strictly on its arguments, without dereferencing any
1029pointer arguments or otherwise accessing any mutable state (e.g. memory, control
1030registers, etc) visible to caller functions. It does not write through any
1031pointer arguments (including <tt><a href="#byval">byval</a></tt> arguments) and
1032never changes any state visible to callers.</dd>
Devang Patel008cd3e2008-09-26 23:51:19 +0000[diff] [blame]1033
Duncan Sands7d94e5a2008-10-06 08:14:18 +0000[diff] [blame]1034<dt><tt><a name="readonly">readonly</a></tt></dt>
1035<dd>This attribute indicates that the function does not write through any
1036pointer arguments (including <tt><a href="#byval">byval</a></tt> arguments)
1037or otherwise modify any state (e.g. memory, control registers, etc) visible to
1038caller functions. It may dereference pointer arguments and read state that may
1039be set in the caller. A readonly function always returns the same value (or
1040throws the same exception) when called with the same set of arguments and global
1041state.</dd>
Bill Wendlingdfaabba2008-11-13 01:02:51 +0000[diff] [blame]1042
1043<dt><tt><a name="ssp">ssp</a></tt></dt>
Bill Wendling34f70d82008-11-26 19:19:05 +0000[diff] [blame]1044<dd>This attribute indicates that the function should emit a stack smashing
Bill Wendlingdfaabba2008-11-13 01:02:51 +0000[diff] [blame]1045protector. It is in the form of a "canary"&mdash;a random value placed on the
1046stack before the local variables that's checked upon return from the function to
1047see if it has been overwritten. A heuristic is used to determine if a function
Bill Wendling34f70d82008-11-26 19:19:05 +0000[diff] [blame]1048needs stack protectors or not.
Bill Wendlingdfaabba2008-11-13 01:02:51 +0000[diff] [blame]1049
Bill Wendlingbe9ec3f2008-11-26 19:07:40 +0000[diff] [blame]1050<p>If a function that has an <tt>ssp</tt> attribute is inlined into a function
1051that doesn't have an <tt>ssp</tt> attribute, then the resulting function will
1052have an <tt>ssp</tt> attribute.</p></dd>
1053
1054<dt><tt>sspreq</tt></dt>
Bill Wendling34f70d82008-11-26 19:19:05 +0000[diff] [blame]1055<dd>This attribute indicates that the function should <em>always</em> emit a
Bill Wendlingdfaabba2008-11-13 01:02:51 +0000[diff] [blame]1056stack smashing protector. This overrides the <tt><a href="#ssp">ssp</a></tt>
Bill Wendling34f70d82008-11-26 19:19:05 +0000[diff] [blame]1057function attribute.
Bill Wendlingbe9ec3f2008-11-26 19:07:40 +0000[diff] [blame]1058
1059<p>If a function that has an <tt>sspreq</tt> attribute is inlined into a
1060function that doesn't have an <tt>sspreq</tt> attribute or which has
1061an <tt>ssp</tt> attribute, then the resulting function will have
1062an <tt>sspreq</tt> attribute.</p></dd>
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1063</dl>
1064
Devang Pateld468f1c2008-09-04 23:05:13 +0000[diff] [blame]1065</div>
1066
1067<!-- ======================================================================= -->
1068<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1069 <a name="moduleasm">Module-Level Inline Assembly</a>
1070</div>
1071
1072<div class="doc_text">
1073<p>
1074Modules may contain "module-level inline asm" blocks, which corresponds to the
1075GCC "file scope inline asm" blocks. These blocks are internally concatenated by
1076LLVM and treated as a single unit, but may be separated in the .ll file if
1077desired. The syntax is very simple:
1078</p>
1079
1080<div class="doc_code">
1081<pre>
1082module asm "inline asm code goes here"
1083module asm "more can go here"
1084</pre>
1085</div>
1086
1087<p>The strings can contain any character by escaping non-printable characters.
1088 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
1089 for the number.
1090</p>
1091
1092<p>
1093 The inline asm code is simply printed to the machine code .s file when
1094 assembly code is generated.
1095</p>
1096</div>
1097
1098<!-- ======================================================================= -->
1099<div class="doc_subsection">
1100 <a name="datalayout">Data Layout</a>
1101</div>
1102
1103<div class="doc_text">
1104<p>A module may specify a target specific data layout string that specifies how
1105data is to be laid out in memory. The syntax for the data layout is simply:</p>
1106<pre> target datalayout = "<i>layout specification</i>"</pre>
1107<p>The <i>layout specification</i> consists of a list of specifications
1108separated by the minus sign character ('-'). Each specification starts with a
1109letter and may include other information after the letter to define some
1110aspect of the data layout. The specifications accepted are as follows: </p>
1111<dl>
1112 <dt><tt>E</tt></dt>
1113 <dd>Specifies that the target lays out data in big-endian form. That is, the
1114 bits with the most significance have the lowest address location.</dd>
1115 <dt><tt>e</tt></dt>
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]1116 <dd>Specifies that the target lays out data in little-endian form. That is,
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1117 the bits with the least significance have the lowest address location.</dd>
1118 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1119 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
1120 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
1121 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
1122 too.</dd>
1123 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1124 <dd>This specifies the alignment for an integer type of a given bit
1125 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1126 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1127 <dd>This specifies the alignment for a vector type of a given bit
1128 <i>size</i>.</dd>
1129 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1130 <dd>This specifies the alignment for a floating point type of a given bit
1131 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
1132 (double).</dd>
1133 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1134 <dd>This specifies the alignment for an aggregate type of a given bit
1135 <i>size</i>.</dd>
1136</dl>
1137<p>When constructing the data layout for a given target, LLVM starts with a
1138default set of specifications which are then (possibly) overriden by the
1139specifications in the <tt>datalayout</tt> keyword. The default specifications
1140are given in this list:</p>
1141<ul>
1142 <li><tt>E</tt> - big endian</li>
1143 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
1144 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1145 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1146 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1147 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]1148 <li><tt>i64:32:64</tt> - i64 has ABI alignment of 32-bits but preferred
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1149 alignment of 64-bits</li>
1150 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1151 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1152 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1153 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1154 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
1155</ul>
Chris Lattner6a9f3c42008-08-05 18:21:08 +0000[diff] [blame]1156<p>When LLVM is determining the alignment for a given type, it uses the
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]1157following rules:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1158<ol>
1159 <li>If the type sought is an exact match for one of the specifications, that
1160 specification is used.</li>
1161 <li>If no match is found, and the type sought is an integer type, then the
1162 smallest integer type that is larger than the bitwidth of the sought type is
1163 used. If none of the specifications are larger than the bitwidth then the the
1164 largest integer type is used. For example, given the default specifications
1165 above, the i7 type will use the alignment of i8 (next largest) while both
1166 i65 and i256 will use the alignment of i64 (largest specified).</li>
1167 <li>If no match is found, and the type sought is a vector type, then the
1168 largest vector type that is smaller than the sought vector type will be used
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]1169 as a fall back. This happens because &lt;128 x double&gt; can be implemented
1170 in terms of 64 &lt;2 x double&gt;, for example.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1171</ol>
1172</div>
1173
1174<!-- *********************************************************************** -->
1175<div class="doc_section"> <a name="typesystem">Type System</a> </div>
1176<!-- *********************************************************************** -->
1177
1178<div class="doc_text">
1179
1180<p>The LLVM type system is one of the most important features of the
1181intermediate representation. Being typed enables a number of
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]1182optimizations to be performed on the intermediate representation directly,
1183without having to do
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1184extra analyses on the side before the transformation. A strong type
1185system makes it easier to read the generated code and enables novel
1186analyses and transformations that are not feasible to perform on normal
1187three address code representations.</p>
1188
1189</div>
1190
1191<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1192<div class="doc_subsection"> <a name="t_classifications">Type
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1193Classifications</a> </div>
1194<div class="doc_text">
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1195<p>The types fall into a few useful
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1196classifications:</p>
1197
1198<table border="1" cellspacing="0" cellpadding="4">
1199 <tbody>
1200 <tr><th>Classification</th><th>Types</th></tr>
1201 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1202 <td><a href="#t_integer">integer</a></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1203 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
1204 </tr>
1205 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1206 <td><a href="#t_floating">floating point</a></td>
1207 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1208 </tr>
1209 <tr>
1210 <td><a name="t_firstclass">first class</a></td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1211 <td><a href="#t_integer">integer</a>,
1212 <a href="#t_floating">floating point</a>,
1213 <a href="#t_pointer">pointer</a>,
Dan Gohmanf6237db2008-06-18 18:42:13 +0000[diff] [blame]1214 <a href="#t_vector">vector</a>,
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]1215 <a href="#t_struct">structure</a>,
1216 <a href="#t_array">array</a>,
Dan Gohmand13951e2008-05-23 22:50:26 +0000[diff] [blame]1217 <a href="#t_label">label</a>.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1218 </td>
1219 </tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1220 <tr>
1221 <td><a href="#t_primitive">primitive</a></td>
1222 <td><a href="#t_label">label</a>,
1223 <a href="#t_void">void</a>,
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1224 <a href="#t_floating">floating point</a>.</td>
1225 </tr>
1226 <tr>
1227 <td><a href="#t_derived">derived</a></td>
1228 <td><a href="#t_integer">integer</a>,
1229 <a href="#t_array">array</a>,
1230 <a href="#t_function">function</a>,
1231 <a href="#t_pointer">pointer</a>,
1232 <a href="#t_struct">structure</a>,
1233 <a href="#t_pstruct">packed structure</a>,
1234 <a href="#t_vector">vector</a>,
1235 <a href="#t_opaque">opaque</a>.
Dan Gohman032ba852008-10-14 16:32:04 +0000[diff] [blame]1236 </td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1237 </tr>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1238 </tbody>
1239</table>
1240
1241<p>The <a href="#t_firstclass">first class</a> types are perhaps the
1242most important. Values of these types are the only ones which can be
1243produced by instructions, passed as arguments, or used as operands to
Dan Gohman4f29e422008-05-23 21:53:15 +0000[diff] [blame]1244instructions.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1245</div>
1246
1247<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1248<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame]1249
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1250<div class="doc_text">
1251<p>The primitive types are the fundamental building blocks of the LLVM
1252system.</p>
1253
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame]1254</div>
1255
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1256<!-- _______________________________________________________________________ -->
1257<div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1258
1259<div class="doc_text">
1260 <table>
1261 <tbody>
1262 <tr><th>Type</th><th>Description</th></tr>
1263 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1264 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1265 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1266 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1267 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1268 </tbody>
1269 </table>
1270</div>
1271
1272<!-- _______________________________________________________________________ -->
1273<div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1274
1275<div class="doc_text">
1276<h5>Overview:</h5>
1277<p>The void type does not represent any value and has no size.</p>
1278
1279<h5>Syntax:</h5>
1280
1281<pre>
1282 void
1283</pre>
1284</div>
1285
1286<!-- _______________________________________________________________________ -->
1287<div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1288
1289<div class="doc_text">
1290<h5>Overview:</h5>
1291<p>The label type represents code labels.</p>
1292
1293<h5>Syntax:</h5>
1294
1295<pre>
1296 label
1297</pre>
1298</div>
1299
1300
1301<!-- ======================================================================= -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1302<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
1303
1304<div class="doc_text">
1305
1306<p>The real power in LLVM comes from the derived types in the system.
1307This is what allows a programmer to represent arrays, functions,
1308pointers, and other useful types. Note that these derived types may be
1309recursive: For example, it is possible to have a two dimensional array.</p>
1310
1311</div>
1312
1313<!-- _______________________________________________________________________ -->
1314<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1315
1316<div class="doc_text">
1317
1318<h5>Overview:</h5>
1319<p>The integer type is a very simple derived type that simply specifies an
1320arbitrary bit width for the integer type desired. Any bit width from 1 bit to
13212^23-1 (about 8 million) can be specified.</p>
1322
1323<h5>Syntax:</h5>
1324
1325<pre>
1326 iN
1327</pre>
1328
1329<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1330value.</p>
1331
1332<h5>Examples:</h5>
1333<table class="layout">
Chris Lattner251ab812007-12-18 06:18:21 +0000[diff] [blame]1334 <tbody>
1335 <tr>
1336 <td><tt>i1</tt></td>
1337 <td>a single-bit integer.</td>
1338 </tr><tr>
1339 <td><tt>i32</tt></td>
1340 <td>a 32-bit integer.</td>
1341 </tr><tr>
1342 <td><tt>i1942652</tt></td>
1343 <td>a really big integer of over 1 million bits.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1344 </tr>
Chris Lattner251ab812007-12-18 06:18:21 +0000[diff] [blame]1345 </tbody>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1346</table>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1347
1348<p>Note that the code generator does not yet support large integer types
1349to be used as function return types. The specific limit on how large a
1350return type the code generator can currently handle is target-dependent;
1351currently it's often 64 bits for 32-bit targets and 128 bits for 64-bit
1352targets.</p>
1353
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1354</div>
1355
1356<!-- _______________________________________________________________________ -->
1357<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
1358
1359<div class="doc_text">
1360
1361<h5>Overview:</h5>
1362
1363<p>The array type is a very simple derived type that arranges elements
1364sequentially in memory. The array type requires a size (number of
1365elements) and an underlying data type.</p>
1366
1367<h5>Syntax:</h5>
1368
1369<pre>
1370 [&lt;# elements&gt; x &lt;elementtype&gt;]
1371</pre>
1372
1373<p>The number of elements is a constant integer value; elementtype may
1374be any type with a size.</p>
1375
1376<h5>Examples:</h5>
1377<table class="layout">
1378 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1379 <td class="left"><tt>[40 x i32]</tt></td>
1380 <td class="left">Array of 40 32-bit integer values.</td>
1381 </tr>
1382 <tr class="layout">
1383 <td class="left"><tt>[41 x i32]</tt></td>
1384 <td class="left">Array of 41 32-bit integer values.</td>
1385 </tr>
1386 <tr class="layout">
1387 <td class="left"><tt>[4 x i8]</tt></td>
1388 <td class="left">Array of 4 8-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1389 </tr>
1390</table>
1391<p>Here are some examples of multidimensional arrays:</p>
1392<table class="layout">
1393 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1394 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1395 <td class="left">3x4 array of 32-bit integer values.</td>
1396 </tr>
1397 <tr class="layout">
1398 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1399 <td class="left">12x10 array of single precision floating point values.</td>
1400 </tr>
1401 <tr class="layout">
1402 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1403 <td class="left">2x3x4 array of 16-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1404 </tr>
1405</table>
1406
1407<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1408length array. Normally, accesses past the end of an array are undefined in
1409LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1410As a special case, however, zero length arrays are recognized to be variable
1411length. This allows implementation of 'pascal style arrays' with the LLVM
1412type "{ i32, [0 x float]}", for example.</p>
1413
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1414<p>Note that the code generator does not yet support large aggregate types
1415to be used as function return types. The specific limit on how large an
1416aggregate return type the code generator can currently handle is
1417target-dependent, and also dependent on the aggregate element types.</p>
1418
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1419</div>
1420
1421<!-- _______________________________________________________________________ -->
1422<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
1423<div class="doc_text">
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1424
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1425<h5>Overview:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1426
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1427<p>The function type can be thought of as a function signature. It
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1428consists of a return type and a list of formal parameter types. The
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1429return type of a function type is a scalar type, a void type, or a struct type.
Devang Pateld5404c02008-03-24 20:52:42 +0000[diff] [blame]1430If the return type is a struct type then all struct elements must be of first
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1431class types, and the struct must have at least one element.</p>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1432
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1433<h5>Syntax:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1434
1435<pre>
1436 &lt;returntype list&gt; (&lt;parameter list&gt;)
1437</pre>
1438
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1439<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
1440specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1441which indicates that the function takes a variable number of arguments.
1442Variable argument functions can access their arguments with the <a
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1443 href="#int_varargs">variable argument handling intrinsic</a> functions.
1444'<tt>&lt;returntype list&gt;</tt>' is a comma-separated list of
1445<a href="#t_firstclass">first class</a> type specifiers.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1446
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1447<h5>Examples:</h5>
1448<table class="layout">
1449 <tr class="layout">
1450 <td class="left"><tt>i32 (i32)</tt></td>
1451 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
1452 </td>
1453 </tr><tr class="layout">
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]1454 <td class="left"><tt>float&nbsp;(i16&nbsp;signext,&nbsp;i32&nbsp;*)&nbsp;*
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1455 </tt></td>
1456 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1457 an <tt>i16</tt> that should be sign extended and a
1458 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
1459 <tt>float</tt>.
1460 </td>
1461 </tr><tr class="layout">
1462 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1463 <td class="left">A vararg function that takes at least one
1464 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1465 which returns an integer. This is the signature for <tt>printf</tt> in
1466 LLVM.
1467 </td>
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1468 </tr><tr class="layout">
1469 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Misha Brukmanafc88b02008-11-27 06:41:20 +0000[diff] [blame]1470 <td class="left">A function taking an <tt>i32</tt>, returning two
1471 <tt>i32</tt> values as an aggregate of type <tt>{ i32, i32 }</tt>
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1472 </td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1473 </tr>
1474</table>
1475
1476</div>
1477<!-- _______________________________________________________________________ -->
1478<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
1479<div class="doc_text">
1480<h5>Overview:</h5>
1481<p>The structure type is used to represent a collection of data members
1482together in memory. The packing of the field types is defined to match
1483the ABI of the underlying processor. The elements of a structure may
1484be any type that has a size.</p>
1485<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1486and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1487field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1488instruction.</p>
1489<h5>Syntax:</h5>
1490<pre> { &lt;type list&gt; }<br></pre>
1491<h5>Examples:</h5>
1492<table class="layout">
1493 <tr class="layout">
1494 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1495 <td class="left">A triple of three <tt>i32</tt> values</td>
1496 </tr><tr class="layout">
1497 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1498 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1499 second element is a <a href="#t_pointer">pointer</a> to a
1500 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1501 an <tt>i32</tt>.</td>
1502 </tr>
1503</table>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1504
1505<p>Note that the code generator does not yet support large aggregate types
1506to be used as function return types. The specific limit on how large an
1507aggregate return type the code generator can currently handle is
1508target-dependent, and also dependent on the aggregate element types.</p>
1509
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1510</div>
1511
1512<!-- _______________________________________________________________________ -->
1513<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1514</div>
1515<div class="doc_text">
1516<h5>Overview:</h5>
1517<p>The packed structure type is used to represent a collection of data members
1518together in memory. There is no padding between fields. Further, the alignment
1519of a packed structure is 1 byte. The elements of a packed structure may
1520be any type that has a size.</p>
1521<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1522and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1523field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1524instruction.</p>
1525<h5>Syntax:</h5>
1526<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1527<h5>Examples:</h5>
1528<table class="layout">
1529 <tr class="layout">
1530 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1531 <td class="left">A triple of three <tt>i32</tt> values</td>
1532 </tr><tr class="layout">
Bill Wendling74d3eac2008-09-07 10:26:33 +0000[diff] [blame]1533 <td class="left">
1534<tt>&lt;&nbsp;{&nbsp;float,&nbsp;i32&nbsp;(i32)*&nbsp;}&nbsp;&gt;</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1535 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1536 second element is a <a href="#t_pointer">pointer</a> to a
1537 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1538 an <tt>i32</tt>.</td>
1539 </tr>
1540</table>
1541</div>
1542
1543<!-- _______________________________________________________________________ -->
1544<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
1545<div class="doc_text">
1546<h5>Overview:</h5>
1547<p>As in many languages, the pointer type represents a pointer or
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]1548reference to another object, which must live in memory. Pointer types may have
1549an optional address space attribute defining the target-specific numbered
1550address space where the pointed-to object resides. The default address space is
1551zero.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1552<h5>Syntax:</h5>
1553<pre> &lt;type&gt; *<br></pre>
1554<h5>Examples:</h5>
1555<table class="layout">
1556 <tr class="layout">
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]1557 <td class="left"><tt>[4 x i32]*</tt></td>
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1558 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1559 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1560 </tr>
1561 <tr class="layout">
1562 <td class="left"><tt>i32 (i32 *) *</tt></td>
1563 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1564 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1565 <tt>i32</tt>.</td>
1566 </tr>
1567 <tr class="layout">
1568 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1569 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1570 that resides in address space #5.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1571 </tr>
1572</table>
1573</div>
1574
1575<!-- _______________________________________________________________________ -->
1576<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
1577<div class="doc_text">
1578
1579<h5>Overview:</h5>
1580
1581<p>A vector type is a simple derived type that represents a vector
1582of elements. Vector types are used when multiple primitive data
1583are operated in parallel using a single instruction (SIMD).
1584A vector type requires a size (number of
1585elements) and an underlying primitive data type. Vectors must have a power
1586of two length (1, 2, 4, 8, 16 ...). Vector types are
1587considered <a href="#t_firstclass">first class</a>.</p>
1588
1589<h5>Syntax:</h5>
1590
1591<pre>
1592 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1593</pre>
1594
1595<p>The number of elements is a constant integer value; elementtype may
1596be any integer or floating point type.</p>
1597
1598<h5>Examples:</h5>
1599
1600<table class="layout">
1601 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1602 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1603 <td class="left">Vector of 4 32-bit integer values.</td>
1604 </tr>
1605 <tr class="layout">
1606 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1607 <td class="left">Vector of 8 32-bit floating-point values.</td>
1608 </tr>
1609 <tr class="layout">
1610 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1611 <td class="left">Vector of 2 64-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1612 </tr>
1613</table>
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]1614
1615<p>Note that the code generator does not yet support large vector types
1616to be used as function return types. The specific limit on how large a
1617vector return type codegen can currently handle is target-dependent;
1618currently it's often a few times longer than a hardware vector register.</p>
1619
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1620</div>
1621
1622<!-- _______________________________________________________________________ -->
1623<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1624<div class="doc_text">
1625
1626<h5>Overview:</h5>
1627
1628<p>Opaque types are used to represent unknown types in the system. This
Gordon Henriksenda0706e2007-10-14 00:34:53 +0000[diff] [blame]1629corresponds (for example) to the C notion of a forward declared structure type.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1630In LLVM, opaque types can eventually be resolved to any type (not just a
1631structure type).</p>
1632
1633<h5>Syntax:</h5>
1634
1635<pre>
1636 opaque
1637</pre>
1638
1639<h5>Examples:</h5>
1640
1641<table class="layout">
1642 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1643 <td class="left"><tt>opaque</tt></td>
1644 <td class="left">An opaque type.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1645 </tr>
1646</table>
1647</div>
1648
1649
1650<!-- *********************************************************************** -->
1651<div class="doc_section"> <a name="constants">Constants</a> </div>
1652<!-- *********************************************************************** -->
1653
1654<div class="doc_text">
1655
1656<p>LLVM has several different basic types of constants. This section describes
1657them all and their syntax.</p>
1658
1659</div>
1660
1661<!-- ======================================================================= -->
1662<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
1663
1664<div class="doc_text">
1665
1666<dl>
1667 <dt><b>Boolean constants</b></dt>
1668
1669 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
1670 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
1671 </dd>
1672
1673 <dt><b>Integer constants</b></dt>
1674
1675 <dd>Standard integers (such as '4') are constants of the <a
1676 href="#t_integer">integer</a> type. Negative numbers may be used with
1677 integer types.
1678 </dd>
1679
1680 <dt><b>Floating point constants</b></dt>
1681
1682 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1683 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]1684 notation (see below). The assembler requires the exact decimal value of
1685 a floating-point constant. For example, the assembler accepts 1.25 but
1686 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
1687 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1688
1689 <dt><b>Null pointer constants</b></dt>
1690
1691 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
1692 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1693
1694</dl>
1695
1696<p>The one non-intuitive notation for constants is the optional hexadecimal form
1697of floating point constants. For example, the form '<tt>double
16980x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
16994.5e+15</tt>'. The only time hexadecimal floating point constants are required
1700(and the only time that they are generated by the disassembler) is when a
1701floating point constant must be emitted but it cannot be represented as a
1702decimal floating point number. For example, NaN's, infinities, and other
1703special values are represented in their IEEE hexadecimal format so that
1704assembly and disassembly do not cause any bits to change in the constants.</p>
1705
1706</div>
1707
1708<!-- ======================================================================= -->
1709<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1710</div>
1711
1712<div class="doc_text">
1713<p>Aggregate constants arise from aggregation of simple constants
1714and smaller aggregate constants.</p>
1715
1716<dl>
1717 <dt><b>Structure constants</b></dt>
1718
1719 <dd>Structure constants are represented with notation similar to structure
1720 type definitions (a comma separated list of elements, surrounded by braces
Chris Lattner6c8de962007-12-25 20:34:52 +0000[diff] [blame]1721 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
1722 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>". Structure constants
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1723 must have <a href="#t_struct">structure type</a>, and the number and
1724 types of elements must match those specified by the type.
1725 </dd>
1726
1727 <dt><b>Array constants</b></dt>
1728
1729 <dd>Array constants are represented with notation similar to array type
1730 definitions (a comma separated list of elements, surrounded by square brackets
1731 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
1732 constants must have <a href="#t_array">array type</a>, and the number and
1733 types of elements must match those specified by the type.
1734 </dd>
1735
1736 <dt><b>Vector constants</b></dt>
1737
1738 <dd>Vector constants are represented with notation similar to vector type
1739 definitions (a comma separated list of elements, surrounded by
1740 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
1741 i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must have <a
1742 href="#t_vector">vector type</a>, and the number and types of elements must
1743 match those specified by the type.
1744 </dd>
1745
1746 <dt><b>Zero initialization</b></dt>
1747
1748 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1749 value to zero of <em>any</em> type, including scalar and aggregate types.
1750 This is often used to avoid having to print large zero initializers (e.g. for
1751 large arrays) and is always exactly equivalent to using explicit zero
1752 initializers.
1753 </dd>
1754</dl>
1755
1756</div>
1757
1758<!-- ======================================================================= -->
1759<div class="doc_subsection">
1760 <a name="globalconstants">Global Variable and Function Addresses</a>
1761</div>
1762
1763<div class="doc_text">
1764
1765<p>The addresses of <a href="#globalvars">global variables</a> and <a
1766href="#functionstructure">functions</a> are always implicitly valid (link-time)
1767constants. These constants are explicitly referenced when the <a
1768href="#identifiers">identifier for the global</a> is used and always have <a
1769href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1770file:</p>
1771
1772<div class="doc_code">
1773<pre>
1774@X = global i32 17
1775@Y = global i32 42
1776@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
1777</pre>
1778</div>
1779
1780</div>
1781
1782<!-- ======================================================================= -->
1783<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
1784<div class="doc_text">
1785 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
1786 no specific value. Undefined values may be of any type and be used anywhere
1787 a constant is permitted.</p>
1788
1789 <p>Undefined values indicate to the compiler that the program is well defined
1790 no matter what value is used, giving the compiler more freedom to optimize.
1791 </p>
1792</div>
1793
1794<!-- ======================================================================= -->
1795<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1796</div>
1797
1798<div class="doc_text">
1799
1800<p>Constant expressions are used to allow expressions involving other constants
1801to be used as constants. Constant expressions may be of any <a
1802href="#t_firstclass">first class</a> type and may involve any LLVM operation
1803that does not have side effects (e.g. load and call are not supported). The
1804following is the syntax for constant expressions:</p>
1805
1806<dl>
1807 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1808 <dd>Truncate a constant to another type. The bit size of CST must be larger
1809 than the bit size of TYPE. Both types must be integers.</dd>
1810
1811 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1812 <dd>Zero extend a constant to another type. The bit size of CST must be
1813 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1814
1815 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1816 <dd>Sign extend a constant to another type. The bit size of CST must be
1817 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1818
1819 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1820 <dd>Truncate a floating point constant to another floating point type. The
1821 size of CST must be larger than the size of TYPE. Both types must be
1822 floating point.</dd>
1823
1824 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1825 <dd>Floating point extend a constant to another type. The size of CST must be
1826 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1827
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]1828 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1829 <dd>Convert a floating point constant to the corresponding unsigned integer
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1830 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1831 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1832 of the same number of elements. If the value won't fit in the integer type,
1833 the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1834
1835 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
1836 <dd>Convert a floating point constant to the corresponding signed integer
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1837 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1838 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1839 of the same number of elements. If the value won't fit in the integer type,
1840 the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1841
1842 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
1843 <dd>Convert an unsigned integer constant to the corresponding floating point
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1844 constant. TYPE must be a scalar or vector floating point type. CST must be of
1845 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1846 of the same number of elements. If the value won't fit in the floating point
1847 type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1848
1849 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
1850 <dd>Convert a signed integer constant to the corresponding floating point
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1851 constant. TYPE must be a scalar or vector floating point type. CST must be of
1852 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1853 of the same number of elements. If the value won't fit in the floating point
1854 type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1855
1856 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1857 <dd>Convert a pointer typed constant to the corresponding integer constant
1858 TYPE must be an integer type. CST must be of pointer type. The CST value is
1859 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1860
1861 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1862 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1863 pointer type. CST must be of integer type. The CST value is zero extended,
1864 truncated, or unchanged to make it fit in a pointer size. This one is
1865 <i>really</i> dangerous!</dd>
1866
1867 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
1868 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1869 identical (same number of bits). The conversion is done as if the CST value
1870 was stored to memory and read back as TYPE. In other words, no bits change
1871 with this operator, just the type. This can be used for conversion of
1872 vector types to any other type, as long as they have the same bit width. For
Dan Gohman7305fa02008-09-08 16:45:59 +0000[diff] [blame]1873 pointers it is only valid to cast to another pointer type. It is not valid
1874 to bitcast to or from an aggregate type.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1875 </dd>
1876
1877 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1878
1879 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1880 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1881 instruction, the index list may have zero or more indexes, which are required
1882 to make sense for the type of "CSTPTR".</dd>
1883
1884 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1885
1886 <dd>Perform the <a href="#i_select">select operation</a> on
1887 constants.</dd>
1888
1889 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1890 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1891
1892 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1893 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
1894
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]1895 <dt><b><tt>vicmp COND ( VAL1, VAL2 )</tt></b></dt>
1896 <dd>Performs the <a href="#i_vicmp">vicmp operation</a> on constants.</dd>
1897
1898 <dt><b><tt>vfcmp COND ( VAL1, VAL2 )</tt></b></dt>
1899 <dd>Performs the <a href="#i_vfcmp">vfcmp operation</a> on constants.</dd>
1900
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1901 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1902
1903 <dd>Perform the <a href="#i_extractelement">extractelement
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]1904 operation</a> on constants.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1905
1906 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1907
1908 <dd>Perform the <a href="#i_insertelement">insertelement
1909 operation</a> on constants.</dd>
1910
1911
1912 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1913
1914 <dd>Perform the <a href="#i_shufflevector">shufflevector
1915 operation</a> on constants.</dd>
1916
1917 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1918
1919 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1920 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
1921 binary</a> operations. The constraints on operands are the same as those for
1922 the corresponding instruction (e.g. no bitwise operations on floating point
1923 values are allowed).</dd>
1924</dl>
1925</div>
1926
1927<!-- *********************************************************************** -->
1928<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1929<!-- *********************************************************************** -->
1930
1931<!-- ======================================================================= -->
1932<div class="doc_subsection">
1933<a name="inlineasm">Inline Assembler Expressions</a>
1934</div>
1935
1936<div class="doc_text">
1937
1938<p>
1939LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1940Module-Level Inline Assembly</a>) through the use of a special value. This
1941value represents the inline assembler as a string (containing the instructions
1942to emit), a list of operand constraints (stored as a string), and a flag that
1943indicates whether or not the inline asm expression has side effects. An example
1944inline assembler expression is:
1945</p>
1946
1947<div class="doc_code">
1948<pre>
1949i32 (i32) asm "bswap $0", "=r,r"
1950</pre>
1951</div>
1952
1953<p>
1954Inline assembler expressions may <b>only</b> be used as the callee operand of
1955a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1956</p>
1957
1958<div class="doc_code">
1959<pre>
1960%X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
1961</pre>
1962</div>
1963
1964<p>
1965Inline asms with side effects not visible in the constraint list must be marked
1966as having side effects. This is done through the use of the
1967'<tt>sideeffect</tt>' keyword, like so:
1968</p>
1969
1970<div class="doc_code">
1971<pre>
1972call void asm sideeffect "eieio", ""()
1973</pre>
1974</div>
1975
1976<p>TODO: The format of the asm and constraints string still need to be
1977documented here. Constraints on what can be done (e.g. duplication, moving, etc
Chris Lattner1cdd64e2008-10-04 18:36:02 +0000[diff] [blame]1978need to be documented). This is probably best done by reference to another
1979document that covers inline asm from a holistic perspective.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1980</p>
1981
1982</div>
1983
1984<!-- *********************************************************************** -->
1985<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1986<!-- *********************************************************************** -->
1987
1988<div class="doc_text">
1989
1990<p>The LLVM instruction set consists of several different
1991classifications of instructions: <a href="#terminators">terminator
1992instructions</a>, <a href="#binaryops">binary instructions</a>,
1993<a href="#bitwiseops">bitwise binary instructions</a>, <a
1994 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1995instructions</a>.</p>
1996
1997</div>
1998
1999<!-- ======================================================================= -->
2000<div class="doc_subsection"> <a name="terminators">Terminator
2001Instructions</a> </div>
2002
2003<div class="doc_text">
2004
2005<p>As mentioned <a href="#functionstructure">previously</a>, every
2006basic block in a program ends with a "Terminator" instruction, which
2007indicates which block should be executed after the current block is
2008finished. These terminator instructions typically yield a '<tt>void</tt>'
2009value: they produce control flow, not values (the one exception being
2010the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
2011<p>There are six different terminator instructions: the '<a
2012 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
2013instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
2014the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
2015 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
2016 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
2017
2018</div>
2019
2020<!-- _______________________________________________________________________ -->
2021<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
2022Instruction</a> </div>
2023<div class="doc_text">
2024<h5>Syntax:</h5>
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]2025<pre>
2026 ret &lt;type&gt; &lt;value&gt; <i>; Return a value from a non-void function</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2027 ret void <i>; Return from void function</i>
2028</pre>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2029
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2030<h5>Overview:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2031
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]2032<p>The '<tt>ret</tt>' instruction is used to return control flow (and
2033optionally a value) from a function back to the caller.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2034<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]2035returns a value and then causes control flow, and one that just causes
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2036control flow to occur.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2037
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2038<h5>Arguments:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2039
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]2040<p>The '<tt>ret</tt>' instruction optionally accepts a single argument,
2041the return value. The type of the return value must be a
2042'<a href="#t_firstclass">first class</a>' type.</p>
2043
2044<p>A function is not <a href="#wellformed">well formed</a> if
2045it it has a non-void return type and contains a '<tt>ret</tt>'
2046instruction with no return value or a return value with a type that
2047does not match its type, or if it has a void return type and contains
2048a '<tt>ret</tt>' instruction with a return value.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2049
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2050<h5>Semantics:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2051
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2052<p>When the '<tt>ret</tt>' instruction is executed, control flow
2053returns back to the calling function's context. If the caller is a "<a
2054 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
2055the instruction after the call. If the caller was an "<a
2056 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
2057at the beginning of the "normal" destination block. If the instruction
2058returns a value, that value shall set the call or invoke instruction's
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]2059return value.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2060
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2061<h5>Example:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]2062
2063<pre>
2064 ret i32 5 <i>; Return an integer value of 5</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2065 ret void <i>; Return from a void function</i>
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]2066 ret { i32, i8 } { i32 4, i8 2 } <i>; Return an aggregate of values 4 and 2</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2067</pre>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]2068
djge93155c2009-01-24 15:58:40 +0000[diff] [blame]2069<p>Note that the code generator does not yet fully support large
2070 return values. The specific sizes that are currently supported are
2071 dependent on the target. For integers, on 32-bit targets the limit
2072 is often 64 bits, and on 64-bit targets the limit is often 128 bits.
2073 For aggregate types, the current limits are dependent on the element
2074 types; for example targets are often limited to 2 total integer
2075 elements and 2 total floating-point elements.</p>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]2076
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2077</div>
2078<!-- _______________________________________________________________________ -->
2079<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
2080<div class="doc_text">
2081<h5>Syntax:</h5>
2082<pre> br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
2083</pre>
2084<h5>Overview:</h5>
2085<p>The '<tt>br</tt>' instruction is used to cause control flow to
2086transfer to a different basic block in the current function. There are
2087two forms of this instruction, corresponding to a conditional branch
2088and an unconditional branch.</p>
2089<h5>Arguments:</h5>
2090<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
2091single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
2092unconditional form of the '<tt>br</tt>' instruction takes a single
2093'<tt>label</tt>' value as a target.</p>
2094<h5>Semantics:</h5>
2095<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
2096argument is evaluated. If the value is <tt>true</tt>, control flows
2097to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
2098control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
2099<h5>Example:</h5>
2100<pre>Test:<br> %cond = <a href="#i_icmp">icmp</a> eq, i32 %a, %b<br> br i1 %cond, label %IfEqual, label %IfUnequal<br>IfEqual:<br> <a
2101 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
2102</div>
2103<!-- _______________________________________________________________________ -->
2104<div class="doc_subsubsection">
2105 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
2106</div>
2107
2108<div class="doc_text">
2109<h5>Syntax:</h5>
2110
2111<pre>
2112 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
2113</pre>
2114
2115<h5>Overview:</h5>
2116
2117<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
2118several different places. It is a generalization of the '<tt>br</tt>'
2119instruction, allowing a branch to occur to one of many possible
2120destinations.</p>
2121
2122
2123<h5>Arguments:</h5>
2124
2125<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
2126comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
2127an array of pairs of comparison value constants and '<tt>label</tt>'s. The
2128table is not allowed to contain duplicate constant entries.</p>
2129
2130<h5>Semantics:</h5>
2131
2132<p>The <tt>switch</tt> instruction specifies a table of values and
2133destinations. When the '<tt>switch</tt>' instruction is executed, this
2134table is searched for the given value. If the value is found, control flow is
2135transfered to the corresponding destination; otherwise, control flow is
2136transfered to the default destination.</p>
2137
2138<h5>Implementation:</h5>
2139
2140<p>Depending on properties of the target machine and the particular
2141<tt>switch</tt> instruction, this instruction may be code generated in different
2142ways. For example, it could be generated as a series of chained conditional
2143branches or with a lookup table.</p>
2144
2145<h5>Example:</h5>
2146
2147<pre>
2148 <i>; Emulate a conditional br instruction</i>
2149 %Val = <a href="#i_zext">zext</a> i1 %value to i32
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]2150 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2151
2152 <i>; Emulate an unconditional br instruction</i>
2153 switch i32 0, label %dest [ ]
2154
2155 <i>; Implement a jump table:</i>
Dan Gohman01852382009-01-04 23:44:43 +0000[diff] [blame]2156 switch i32 %val, label %otherwise [ i32 0, label %onzero
2157 i32 1, label %onone
2158 i32 2, label %ontwo ]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2159</pre>
2160</div>
2161
2162<!-- _______________________________________________________________________ -->
2163<div class="doc_subsubsection">
2164 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
2165</div>
2166
2167<div class="doc_text">
2168
2169<h5>Syntax:</h5>
2170
2171<pre>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]2172 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] &lt;ptr to function ty&gt; &lt;function ptr val&gt;(&lt;function args&gt;) [<a href="#fnattrs">fn attrs</a>]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2173 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
2174</pre>
2175
2176<h5>Overview:</h5>
2177
2178<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
2179function, with the possibility of control flow transfer to either the
2180'<tt>normal</tt>' label or the
2181'<tt>exception</tt>' label. If the callee function returns with the
2182"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
2183"normal" label. If the callee (or any indirect callees) returns with the "<a
2184href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]2185continued at the dynamically nearest "exception" label.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2186
2187<h5>Arguments:</h5>
2188
2189<p>This instruction requires several arguments:</p>
2190
2191<ol>
2192 <li>
2193 The optional "cconv" marker indicates which <a href="#callingconv">calling
2194 convention</a> the call should use. If none is specified, the call defaults
2195 to using C calling conventions.
2196 </li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]2197
2198 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
2199 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>',
2200 and '<tt>inreg</tt>' attributes are valid here.</li>
2201
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2202 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
2203 function value being invoked. In most cases, this is a direct function
2204 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
2205 an arbitrary pointer to function value.
2206 </li>
2207
2208 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
2209 function to be invoked. </li>
2210
2211 <li>'<tt>function args</tt>': argument list whose types match the function
2212 signature argument types. If the function signature indicates the function
2213 accepts a variable number of arguments, the extra arguments can be
2214 specified. </li>
2215
2216 <li>'<tt>normal label</tt>': the label reached when the called function
2217 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
2218
2219 <li>'<tt>exception label</tt>': the label reached when a callee returns with
2220 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
2221
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]2222 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]2223 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
2224 '<tt>readnone</tt>' attributes are valid here.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2225</ol>
2226
2227<h5>Semantics:</h5>
2228
2229<p>This instruction is designed to operate as a standard '<tt><a
2230href="#i_call">call</a></tt>' instruction in most regards. The primary
2231difference is that it establishes an association with a label, which is used by
2232the runtime library to unwind the stack.</p>
2233
2234<p>This instruction is used in languages with destructors to ensure that proper
2235cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
2236exception. Additionally, this is important for implementation of
2237'<tt>catch</tt>' clauses in high-level languages that support them.</p>
2238
2239<h5>Example:</h5>
2240<pre>
Nick Lewyckya1c11a12008-03-16 07:18:12 +0000[diff] [blame]2241 %retval = invoke i32 @Test(i32 15) to label %Continue
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2242 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckya1c11a12008-03-16 07:18:12 +0000[diff] [blame]2243 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2244 unwind label %TestCleanup <i>; {i32}:retval set</i>
2245</pre>
2246</div>
2247
2248
2249<!-- _______________________________________________________________________ -->
2250
2251<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
2252Instruction</a> </div>
2253
2254<div class="doc_text">
2255
2256<h5>Syntax:</h5>
2257<pre>
2258 unwind
2259</pre>
2260
2261<h5>Overview:</h5>
2262
2263<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
2264at the first callee in the dynamic call stack which used an <a
2265href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
2266primarily used to implement exception handling.</p>
2267
2268<h5>Semantics:</h5>
2269
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]2270<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2271immediately halt. The dynamic call stack is then searched for the first <a
2272href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
2273execution continues at the "exceptional" destination block specified by the
2274<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
2275dynamic call chain, undefined behavior results.</p>
2276</div>
2277
2278<!-- _______________________________________________________________________ -->
2279
2280<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
2281Instruction</a> </div>
2282
2283<div class="doc_text">
2284
2285<h5>Syntax:</h5>
2286<pre>
2287 unreachable
2288</pre>
2289
2290<h5>Overview:</h5>
2291
2292<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
2293instruction is used to inform the optimizer that a particular portion of the
2294code is not reachable. This can be used to indicate that the code after a
2295no-return function cannot be reached, and other facts.</p>
2296
2297<h5>Semantics:</h5>
2298
2299<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
2300</div>
2301
2302
2303
2304<!-- ======================================================================= -->
2305<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
2306<div class="doc_text">
2307<p>Binary operators are used to do most of the computation in a
Chris Lattnerab596d92008-04-01 18:47:32 +0000[diff] [blame]2308program. They require two operands of the same type, execute an operation on them, and
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2309produce a single value. The operands might represent
2310multiple data, as is the case with the <a href="#t_vector">vector</a> data type.
Chris Lattnerab596d92008-04-01 18:47:32 +0000[diff] [blame]2311The result value has the same type as its operands.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2312<p>There are several different binary operators:</p>
2313</div>
2314<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2315<div class="doc_subsubsection">
2316 <a name="i_add">'<tt>add</tt>' Instruction</a>
2317</div>
2318
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2319<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2320
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2321<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2322
2323<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2324 &lt;result&gt; = add &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2325</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2326
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2327<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2328
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2329<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2330
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2331<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2332
2333<p>The two arguments to the '<tt>add</tt>' instruction must be <a
2334 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>, or
2335 <a href="#t_vector">vector</a> values. Both arguments must have identical
2336 types.</p>
2337
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2338<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2339
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2340<p>The value produced is the integer or floating point sum of the two
2341operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2342
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2343<p>If an integer sum has unsigned overflow, the result returned is the
2344mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2345the result.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2346
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2347<p>Because LLVM integers use a two's complement representation, this
2348instruction is appropriate for both signed and unsigned integers.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2349
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2350<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2351
2352<pre>
2353 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2354</pre>
2355</div>
2356<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2357<div class="doc_subsubsection">
2358 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
2359</div>
2360
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2361<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2362
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2363<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2364
2365<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2366 &lt;result&gt; = sub &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2367</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2368
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2369<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2370
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2371<p>The '<tt>sub</tt>' instruction returns the difference of its two
2372operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2373
2374<p>Note that the '<tt>sub</tt>' instruction is used to represent the
2375'<tt>neg</tt>' instruction present in most other intermediate
2376representations.</p>
2377
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2378<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2379
2380<p>The two arguments to the '<tt>sub</tt>' instruction must be <a
2381 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2382 or <a href="#t_vector">vector</a> values. Both arguments must have identical
2383 types.</p>
2384
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2385<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2386
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2387<p>The value produced is the integer or floating point difference of
2388the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2389
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2390<p>If an integer difference has unsigned overflow, the result returned is the
2391mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2392the result.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2393
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2394<p>Because LLVM integers use a two's complement representation, this
2395instruction is appropriate for both signed and unsigned integers.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2396
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2397<h5>Example:</h5>
2398<pre>
2399 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
2400 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
2401</pre>
2402</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2403
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2404<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2405<div class="doc_subsubsection">
2406 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
2407</div>
2408
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2409<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2410
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2411<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2412<pre> &lt;result&gt; = mul &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2413</pre>
2414<h5>Overview:</h5>
2415<p>The '<tt>mul</tt>' instruction returns the product of its two
2416operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2417
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2418<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2419
2420<p>The two arguments to the '<tt>mul</tt>' instruction must be <a
2421href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2422or <a href="#t_vector">vector</a> values. Both arguments must have identical
2423types.</p>
2424
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2425<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2426
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2427<p>The value produced is the integer or floating point product of the
2428two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2429
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2430<p>If the result of an integer multiplication has unsigned overflow,
2431the result returned is the mathematical result modulo
24322<sup>n</sup>, where n is the bit width of the result.</p>
2433<p>Because LLVM integers use a two's complement representation, and the
2434result is the same width as the operands, this instruction returns the
2435correct result for both signed and unsigned integers. If a full product
2436(e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands
2437should be sign-extended or zero-extended as appropriate to the
2438width of the full product.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2439<h5>Example:</h5>
2440<pre> &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
2441</pre>
2442</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2443
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2444<!-- _______________________________________________________________________ -->
2445<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2446</a></div>
2447<div class="doc_text">
2448<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2449<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2450</pre>
2451<h5>Overview:</h5>
2452<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
2453operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2454
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2455<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2456
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2457<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2458<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2459values. Both arguments must have identical types.</p>
2460
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2461<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2462
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2463<p>The value produced is the unsigned integer quotient of the two operands.</p>
2464<p>Note that unsigned integer division and signed integer division are distinct
2465operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
2466<p>Division by zero leads to undefined behavior.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2467<h5>Example:</h5>
2468<pre> &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
2469</pre>
2470</div>
2471<!-- _______________________________________________________________________ -->
2472<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2473</a> </div>
2474<div class="doc_text">
2475<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2476<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2477 &lt;result&gt; = sdiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2478</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2479
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2480<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2481
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2482<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
2483operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2484
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2485<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2486
2487<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
2488<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2489values. Both arguments must have identical types.</p>
2490
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2491<h5>Semantics:</h5>
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]2492<p>The value produced is the signed integer quotient of the two operands rounded towards zero.</p>
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2493<p>Note that signed integer division and unsigned integer division are distinct
2494operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
2495<p>Division by zero leads to undefined behavior. Overflow also leads to
2496undefined behavior; this is a rare case, but can occur, for example,
2497by doing a 32-bit division of -2147483648 by -1.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2498<h5>Example:</h5>
2499<pre> &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
2500</pre>
2501</div>
2502<!-- _______________________________________________________________________ -->
2503<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
2504Instruction</a> </div>
2505<div class="doc_text">
2506<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2507<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2508 &lt;result&gt; = fdiv &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2509</pre>
2510<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2511
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2512<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
2513operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2514
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2515<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2516
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2517<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2518<a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2519of floating point values. Both arguments must have identical types.</p>
2520
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2521<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2522
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2523<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2524
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2525<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2526
2527<pre>
2528 &lt;result&gt; = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2529</pre>
2530</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2531
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2532<!-- _______________________________________________________________________ -->
2533<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2534</div>
2535<div class="doc_text">
2536<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2537<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2538</pre>
2539<h5>Overview:</h5>
2540<p>The '<tt>urem</tt>' instruction returns the remainder from the
2541unsigned division of its two arguments.</p>
2542<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2543<p>The two arguments to the '<tt>urem</tt>' instruction must be
2544<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2545values. Both arguments must have identical types.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2546<h5>Semantics:</h5>
2547<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]2548This instruction always performs an unsigned division to get the remainder.</p>
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2549<p>Note that unsigned integer remainder and signed integer remainder are
2550distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
2551<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2552<h5>Example:</h5>
2553<pre> &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2554</pre>
2555
2556</div>
2557<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2558<div class="doc_subsubsection">
2559 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
2560</div>
2561
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2562<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2563
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2564<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2565
2566<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2567 &lt;result&gt; = srem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2568</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2569
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2570<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2571
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2572<p>The '<tt>srem</tt>' instruction returns the remainder from the
Dan Gohman3e3fd8c2007-11-05 23:35:22 +0000[diff] [blame]2573signed division of its two operands. This instruction can also take
2574<a href="#t_vector">vector</a> versions of the values in which case
2575the elements must be integers.</p>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]2576
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2577<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2578
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2579<p>The two arguments to the '<tt>srem</tt>' instruction must be
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2580<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2581values. Both arguments must have identical types.</p>
2582
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2583<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2584
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2585<p>This instruction returns the <i>remainder</i> of a division (where the result
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2586has the same sign as the dividend, <tt>op1</tt>), not the <i>modulo</i>
2587operator (where the result has the same sign as the divisor, <tt>op2</tt>) of
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2588a value. For more information about the difference, see <a
2589 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
2590Math Forum</a>. For a table of how this is implemented in various languages,
2591please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
2592Wikipedia: modulo operation</a>.</p>
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2593<p>Note that signed integer remainder and unsigned integer remainder are
2594distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
2595<p>Taking the remainder of a division by zero leads to undefined behavior.
2596Overflow also leads to undefined behavior; this is a rare case, but can occur,
2597for example, by taking the remainder of a 32-bit division of -2147483648 by -1.
2598(The remainder doesn't actually overflow, but this rule lets srem be
2599implemented using instructions that return both the result of the division
2600and the remainder.)</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2601<h5>Example:</h5>
2602<pre> &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2603</pre>
2604
2605</div>
2606<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2607<div class="doc_subsubsection">
2608 <a name="i_frem">'<tt>frem</tt>' Instruction</a> </div>
2609
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2610<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2611
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2612<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2613<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2614</pre>
2615<h5>Overview:</h5>
2616<p>The '<tt>frem</tt>' instruction returns the remainder from the
2617division of its two operands.</p>
2618<h5>Arguments:</h5>
2619<p>The two arguments to the '<tt>frem</tt>' instruction must be
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2620<a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2621of floating point values. Both arguments must have identical types.</p>
2622
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2623<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2624
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]2625<p>This instruction returns the <i>remainder</i> of a division.
2626The remainder has the same sign as the dividend.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2627
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2628<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2629
2630<pre>
2631 &lt;result&gt; = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2632</pre>
2633</div>
2634
2635<!-- ======================================================================= -->
2636<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2637Operations</a> </div>
2638<div class="doc_text">
2639<p>Bitwise binary operators are used to do various forms of
2640bit-twiddling in a program. They are generally very efficient
2641instructions and can commonly be strength reduced from other
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]2642instructions. They require two operands of the same type, execute an operation on them,
2643and produce a single value. The resulting value is the same type as its operands.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2644</div>
2645
2646<!-- _______________________________________________________________________ -->
2647<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2648Instruction</a> </div>
2649<div class="doc_text">
2650<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2651<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2652</pre>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2653
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2654<h5>Overview:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2655
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2656<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2657the left a specified number of bits.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2658
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2659<h5>Arguments:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2660
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2661<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
Nate Begemanbb1ce942008-07-29 15:49:41 +0000[diff] [blame]2662 href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2663type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2664
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2665<h5>Semantics:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2666
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2667<p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod 2<sup>n</sup>,
2668where n is the width of the result. If <tt>op2</tt> is (statically or dynamically) negative or
Mon P Wangb0f51822008-12-10 08:55:09 +0000[diff] [blame]2669equal to or larger than the number of bits in <tt>op1</tt>, the result is undefined.
2670If the arguments are vectors, each vector element of <tt>op1</tt> is shifted by the
2671corresponding shift amount in <tt>op2</tt>.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2672
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2673<h5>Example:</h5><pre>
2674 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2675 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2676 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2677 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Mon P Wang9901e732008-12-09 05:46:39 +0000[diff] [blame]2678 &lt;result&gt; = shl &lt;2 x i32&gt; &lt; i32 1, i32 1&gt;, &lt; i32 1, i32 2&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 2, i32 4&gt;</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2679</pre>
2680</div>
2681<!-- _______________________________________________________________________ -->
2682<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2683Instruction</a> </div>
2684<div class="doc_text">
2685<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2686<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2687</pre>
2688
2689<h5>Overview:</h5>
2690<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2691operand shifted to the right a specified number of bits with zero fill.</p>
2692
2693<h5>Arguments:</h5>
2694<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Nate Begemanbb1ce942008-07-29 15:49:41 +0000[diff] [blame]2695<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2696type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2697
2698<h5>Semantics:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2699
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2700<p>This instruction always performs a logical shift right operation. The most
2701significant bits of the result will be filled with zero bits after the
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2702shift. If <tt>op2</tt> is (statically or dynamically) equal to or larger than
Mon P Wangb0f51822008-12-10 08:55:09 +0000[diff] [blame]2703the number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
2704vectors, each vector element of <tt>op1</tt> is shifted by the corresponding shift
2705amount in <tt>op2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2706
2707<h5>Example:</h5>
2708<pre>
2709 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2710 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2711 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2712 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2713 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Mon P Wang9901e732008-12-09 05:46:39 +0000[diff] [blame]2714 &lt;result&gt; = lshr &lt;2 x i32&gt; &lt; i32 -2, i32 4&gt;, &lt; i32 1, i32 2&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0x7FFFFFFF, i32 1&gt;</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2715</pre>
2716</div>
2717
2718<!-- _______________________________________________________________________ -->
2719<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2720Instruction</a> </div>
2721<div class="doc_text">
2722
2723<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2724<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2725</pre>
2726
2727<h5>Overview:</h5>
2728<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2729operand shifted to the right a specified number of bits with sign extension.</p>
2730
2731<h5>Arguments:</h5>
2732<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Nate Begemanbb1ce942008-07-29 15:49:41 +0000[diff] [blame]2733<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2734type. '<tt>op2</tt>' is treated as an unsigned value.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2735
2736<h5>Semantics:</h5>
2737<p>This instruction always performs an arithmetic shift right operation,
2738The most significant bits of the result will be filled with the sign bit
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2739of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
Mon P Wangb0f51822008-12-10 08:55:09 +0000[diff] [blame]2740larger than the number of bits in <tt>op1</tt>, the result is undefined. If the
2741arguments are vectors, each vector element of <tt>op1</tt> is shifted by the
2742corresponding shift amount in <tt>op2</tt>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2743
2744<h5>Example:</h5>
2745<pre>
2746 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2747 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2748 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2749 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2750 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Mon P Wang9901e732008-12-09 05:46:39 +0000[diff] [blame]2751 &lt;result&gt; = ashr &lt;2 x i32&gt; &lt; i32 -2, i32 4&gt;, &lt; i32 1, i32 3&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 -1, i32 0&gt;</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2752</pre>
2753</div>
2754
2755<!-- _______________________________________________________________________ -->
2756<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2757Instruction</a> </div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2758
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2759<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2760
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2761<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2762
2763<pre>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2764 &lt;result&gt; = and &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2765</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2766
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2767<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2768
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2769<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2770its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2771
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2772<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2773
2774<p>The two arguments to the '<tt>and</tt>' instruction must be
2775<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2776values. Both arguments must have identical types.</p>
2777
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2778<h5>Semantics:</h5>
2779<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
2780<p> </p>
Bill Wendling1e19eac2008-09-07 10:29:20 +0000[diff] [blame]2781<div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2782<table border="1" cellspacing="0" cellpadding="4">
2783 <tbody>
2784 <tr>
2785 <td>In0</td>
2786 <td>In1</td>
2787 <td>Out</td>
2788 </tr>
2789 <tr>
2790 <td>0</td>
2791 <td>0</td>
2792 <td>0</td>
2793 </tr>
2794 <tr>
2795 <td>0</td>
2796 <td>1</td>
2797 <td>0</td>
2798 </tr>
2799 <tr>
2800 <td>1</td>
2801 <td>0</td>
2802 <td>0</td>
2803 </tr>
2804 <tr>
2805 <td>1</td>
2806 <td>1</td>
2807 <td>1</td>
2808 </tr>
2809 </tbody>
2810</table>
2811</div>
2812<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2813<pre>
2814 &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2815 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2816 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
2817</pre>
2818</div>
2819<!-- _______________________________________________________________________ -->
2820<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
2821<div class="doc_text">
2822<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2823<pre> &lt;result&gt; = or &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2824</pre>
2825<h5>Overview:</h5>
2826<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2827or of its two operands.</p>
2828<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2829
2830<p>The two arguments to the '<tt>or</tt>' instruction must be
2831<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2832values. Both arguments must have identical types.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2833<h5>Semantics:</h5>
2834<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
2835<p> </p>
Bill Wendling1e19eac2008-09-07 10:29:20 +0000[diff] [blame]2836<div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2837<table border="1" cellspacing="0" cellpadding="4">
2838 <tbody>
2839 <tr>
2840 <td>In0</td>
2841 <td>In1</td>
2842 <td>Out</td>
2843 </tr>
2844 <tr>
2845 <td>0</td>
2846 <td>0</td>
2847 <td>0</td>
2848 </tr>
2849 <tr>
2850 <td>0</td>
2851 <td>1</td>
2852 <td>1</td>
2853 </tr>
2854 <tr>
2855 <td>1</td>
2856 <td>0</td>
2857 <td>1</td>
2858 </tr>
2859 <tr>
2860 <td>1</td>
2861 <td>1</td>
2862 <td>1</td>
2863 </tr>
2864 </tbody>
2865</table>
2866</div>
2867<h5>Example:</h5>
2868<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2869 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2870 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
2871</pre>
2872</div>
2873<!-- _______________________________________________________________________ -->
2874<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2875Instruction</a> </div>
2876<div class="doc_text">
2877<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]2878<pre> &lt;result&gt; = xor &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2879</pre>
2880<h5>Overview:</h5>
2881<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2882or of its two operands. The <tt>xor</tt> is used to implement the
2883"one's complement" operation, which is the "~" operator in C.</p>
2884<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2885<p>The two arguments to the '<tt>xor</tt>' instruction must be
2886<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2887values. Both arguments must have identical types.</p>
2888
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2889<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2890
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2891<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
2892<p> </p>
Bill Wendling1e19eac2008-09-07 10:29:20 +0000[diff] [blame]2893<div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2894<table border="1" cellspacing="0" cellpadding="4">
2895 <tbody>
2896 <tr>
2897 <td>In0</td>
2898 <td>In1</td>
2899 <td>Out</td>
2900 </tr>
2901 <tr>
2902 <td>0</td>
2903 <td>0</td>
2904 <td>0</td>
2905 </tr>
2906 <tr>
2907 <td>0</td>
2908 <td>1</td>
2909 <td>1</td>
2910 </tr>
2911 <tr>
2912 <td>1</td>
2913 <td>0</td>
2914 <td>1</td>
2915 </tr>
2916 <tr>
2917 <td>1</td>
2918 <td>1</td>
2919 <td>0</td>
2920 </tr>
2921 </tbody>
2922</table>
2923</div>
2924<p> </p>
2925<h5>Example:</h5>
2926<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2927 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2928 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2929 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
2930</pre>
2931</div>
2932
2933<!-- ======================================================================= -->
2934<div class="doc_subsection">
2935 <a name="vectorops">Vector Operations</a>
2936</div>
2937
2938<div class="doc_text">
2939
2940<p>LLVM supports several instructions to represent vector operations in a
2941target-independent manner. These instructions cover the element-access and
2942vector-specific operations needed to process vectors effectively. While LLVM
2943does directly support these vector operations, many sophisticated algorithms
2944will want to use target-specific intrinsics to take full advantage of a specific
2945target.</p>
2946
2947</div>
2948
2949<!-- _______________________________________________________________________ -->
2950<div class="doc_subsubsection">
2951 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2952</div>
2953
2954<div class="doc_text">
2955
2956<h5>Syntax:</h5>
2957
2958<pre>
2959 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
2960</pre>
2961
2962<h5>Overview:</h5>
2963
2964<p>
2965The '<tt>extractelement</tt>' instruction extracts a single scalar
2966element from a vector at a specified index.
2967</p>
2968
2969
2970<h5>Arguments:</h5>
2971
2972<p>
2973The first operand of an '<tt>extractelement</tt>' instruction is a
2974value of <a href="#t_vector">vector</a> type. The second operand is
2975an index indicating the position from which to extract the element.
2976The index may be a variable.</p>
2977
2978<h5>Semantics:</h5>
2979
2980<p>
2981The result is a scalar of the same type as the element type of
2982<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2983<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2984results are undefined.
2985</p>
2986
2987<h5>Example:</h5>
2988
2989<pre>
2990 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
2991</pre>
2992</div>
2993
2994
2995<!-- _______________________________________________________________________ -->
2996<div class="doc_subsubsection">
2997 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2998</div>
2999
3000<div class="doc_text">
3001
3002<h5>Syntax:</h5>
3003
3004<pre>
Dan Gohmanbcc3c502008-05-12 23:38:42 +0000[diff] [blame]3005 &lt;result&gt; = insertelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, &lt;ty&gt; &lt;elt&gt;, i32 &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3006</pre>
3007
3008<h5>Overview:</h5>
3009
3010<p>
3011The '<tt>insertelement</tt>' instruction inserts a scalar
3012element into a vector at a specified index.
3013</p>
3014
3015
3016<h5>Arguments:</h5>
3017
3018<p>
3019The first operand of an '<tt>insertelement</tt>' instruction is a
3020value of <a href="#t_vector">vector</a> type. The second operand is a
3021scalar value whose type must equal the element type of the first
3022operand. The third operand is an index indicating the position at
3023which to insert the value. The index may be a variable.</p>
3024
3025<h5>Semantics:</h5>
3026
3027<p>
3028The result is a vector of the same type as <tt>val</tt>. Its
3029element values are those of <tt>val</tt> except at position
3030<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
3031exceeds the length of <tt>val</tt>, the results are undefined.
3032</p>
3033
3034<h5>Example:</h5>
3035
3036<pre>
3037 %result = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
3038</pre>
3039</div>
3040
3041<!-- _______________________________________________________________________ -->
3042<div class="doc_subsubsection">
3043 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
3044</div>
3045
3046<div class="doc_text">
3047
3048<h5>Syntax:</h5>
3049
3050<pre>
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]3051 &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;m x i32&gt; &lt;mask&gt; <i>; yields &lt;m x &lt;ty&gt;&gt;</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3052</pre>
3053
3054<h5>Overview:</h5>
3055
3056<p>
3057The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]3058from two input vectors, returning a vector with the same element type as
3059the input and length that is the same as the shuffle mask.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3060</p>
3061
3062<h5>Arguments:</h5>
3063
3064<p>
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]3065The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
3066with types that match each other. The third argument is a shuffle mask whose
3067element type is always 'i32'. The result of the instruction is a vector whose
3068length is the same as the shuffle mask and whose element type is the same as
3069the element type of the first two operands.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3070</p>
3071
3072<p>
3073The shuffle mask operand is required to be a constant vector with either
3074constant integer or undef values.
3075</p>
3076
3077<h5>Semantics:</h5>
3078
3079<p>
3080The elements of the two input vectors are numbered from left to right across
3081both of the vectors. The shuffle mask operand specifies, for each element of
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]3082the result vector, which element of the two input vectors the result element
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3083gets. The element selector may be undef (meaning "don't care") and the second
3084operand may be undef if performing a shuffle from only one vector.
3085</p>
3086
3087<h5>Example:</h5>
3088
3089<pre>
3090 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
3091 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
3092 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
3093 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i> - Identity shuffle.
Mon P Wangbff5d9c2008-11-10 04:46:22 +0000[diff] [blame]3094 %result = shufflevector &lt;8 x i32&gt; %v1, &lt;8 x i32&gt; undef,
3095 &lt;4 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3&gt; <i>; yields &lt;4 x i32&gt;</i>
3096 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
3097 &lt;8 x i32&gt; &lt;i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7 &gt; <i>; yields &lt;8 x i32&gt;</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3098</pre>
3099</div>
3100
3101
3102<!-- ======================================================================= -->
3103<div class="doc_subsection">
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3104 <a name="aggregateops">Aggregate Operations</a>
3105</div>
3106
3107<div class="doc_text">
3108
3109<p>LLVM supports several instructions for working with aggregate values.
3110</p>
3111
3112</div>
3113
3114<!-- _______________________________________________________________________ -->
3115<div class="doc_subsubsection">
3116 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
3117</div>
3118
3119<div class="doc_text">
3120
3121<h5>Syntax:</h5>
3122
3123<pre>
3124 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
3125</pre>
3126
3127<h5>Overview:</h5>
3128
3129<p>
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]3130The '<tt>extractvalue</tt>' instruction extracts the value of a struct field
3131or array element from an aggregate value.
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3132</p>
3133
3134
3135<h5>Arguments:</h5>
3136
3137<p>
3138The first operand of an '<tt>extractvalue</tt>' instruction is a
3139value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a>
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]3140type. The operands are constant indices to specify which value to extract
Dan Gohmane5febe42008-05-31 00:58:22 +0000[diff] [blame]3141in a similar manner as indices in a
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3142'<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
3143</p>
3144
3145<h5>Semantics:</h5>
3146
3147<p>
3148The result is the value at the position in the aggregate specified by
3149the index operands.
3150</p>
3151
3152<h5>Example:</h5>
3153
3154<pre>
Dan Gohmane5febe42008-05-31 00:58:22 +0000[diff] [blame]3155 %result = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3156</pre>
3157</div>
3158
3159
3160<!-- _______________________________________________________________________ -->
3161<div class="doc_subsubsection">
3162 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
3163</div>
3164
3165<div class="doc_text">
3166
3167<h5>Syntax:</h5>
3168
3169<pre>
Dan Gohmane5febe42008-05-31 00:58:22 +0000[diff] [blame]3170 &lt;result&gt; = insertvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;ty&gt; &lt;val&gt;, &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3171</pre>
3172
3173<h5>Overview:</h5>
3174
3175<p>
3176The '<tt>insertvalue</tt>' instruction inserts a value
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]3177into a struct field or array element in an aggregate.
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3178</p>
3179
3180
3181<h5>Arguments:</h5>
3182
3183<p>
3184The first operand of an '<tt>insertvalue</tt>' instruction is a
3185value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a> type.
3186The second operand is a first-class value to insert.
Dan Gohman4f29e422008-05-23 21:53:15 +0000[diff] [blame]3187The following operands are constant indices
Dan Gohmane5febe42008-05-31 00:58:22 +0000[diff] [blame]3188indicating the position at which to insert the value in a similar manner as
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]3189indices in a
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3190'<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
3191The value to insert must have the same type as the value identified
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]3192by the indices.
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]3193</p>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3194
3195<h5>Semantics:</h5>
3196
3197<p>
3198The result is an aggregate of the same type as <tt>val</tt>. Its
3199value is that of <tt>val</tt> except that the value at the position
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]3200specified by the indices is that of <tt>elt</tt>.
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3201</p>
3202
3203<h5>Example:</h5>
3204
3205<pre>
Dan Gohmanb1aab4e2008-06-23 15:26:37 +0000[diff] [blame]3206 %result = insertvalue {i32, float} %agg, i32 1, 0 <i>; yields {i32, float}</i>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]3207</pre>
3208</div>
3209
3210
3211<!-- ======================================================================= -->
3212<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3213 <a name="memoryops">Memory Access and Addressing Operations</a>
3214</div>
3215
3216<div class="doc_text">
3217
3218<p>A key design point of an SSA-based representation is how it
3219represents memory. In LLVM, no memory locations are in SSA form, which
3220makes things very simple. This section describes how to read, write,
3221allocate, and free memory in LLVM.</p>
3222
3223</div>
3224
3225<!-- _______________________________________________________________________ -->
3226<div class="doc_subsubsection">
3227 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
3228</div>
3229
3230<div class="doc_text">
3231
3232<h5>Syntax:</h5>
3233
3234<pre>
3235 &lt;result&gt; = malloc &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
3236</pre>
3237
3238<h5>Overview:</h5>
3239
3240<p>The '<tt>malloc</tt>' instruction allocates memory from the system
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]3241heap and returns a pointer to it. The object is always allocated in the generic
3242address space (address space zero).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3243
3244<h5>Arguments:</h5>
3245
3246<p>The '<tt>malloc</tt>' instruction allocates
3247<tt>sizeof(&lt;type&gt;)*NumElements</tt>
3248bytes of memory from the operating system and returns a pointer of the
3249appropriate type to the program. If "NumElements" is specified, it is the
Gabor Greif5082cf42008-02-09 22:24:34 +0000[diff] [blame]3250number of elements allocated, otherwise "NumElements" is defaulted to be one.
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3251If a constant alignment is specified, the value result of the allocation is guaranteed to
Gabor Greif5082cf42008-02-09 22:24:34 +0000[diff] [blame]3252be aligned to at least that boundary. If not specified, or if zero, the target can
3253choose to align the allocation on any convenient boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3254
3255<p>'<tt>type</tt>' must be a sized type.</p>
3256
3257<h5>Semantics:</h5>
3258
3259<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
Nick Lewyckyff384472008-11-24 03:41:24 +0000[diff] [blame]3260a pointer is returned. The result of a zero byte allocation is undefined. The
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]3261result is null if there is insufficient memory available.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3262
3263<h5>Example:</h5>
3264
3265<pre>
Dan Gohmanf54f50a2009-01-04 23:49:44 +0000[diff] [blame]3266 %array = malloc [4 x i8] <i>; yields {[%4 x i8]*}:array</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3267
3268 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
3269 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
3270 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
3271 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
3272 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
3273</pre>
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]3274
3275<p>Note that the code generator does not yet respect the
3276 alignment value.</p>
3277
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3278</div>
3279
3280<!-- _______________________________________________________________________ -->
3281<div class="doc_subsubsection">
3282 <a name="i_free">'<tt>free</tt>' Instruction</a>
3283</div>
3284
3285<div class="doc_text">
3286
3287<h5>Syntax:</h5>
3288
3289<pre>
Dan Gohmanf54f50a2009-01-04 23:49:44 +0000[diff] [blame]3290 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3291</pre>
3292
3293<h5>Overview:</h5>
3294
3295<p>The '<tt>free</tt>' instruction returns memory back to the unused
3296memory heap to be reallocated in the future.</p>
3297
3298<h5>Arguments:</h5>
3299
3300<p>'<tt>value</tt>' shall be a pointer value that points to a value
3301that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
3302instruction.</p>
3303
3304<h5>Semantics:</h5>
3305
3306<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner329d6532008-04-19 22:41:32 +0000[diff] [blame]3307after this instruction executes. If the pointer is null, the operation
3308is a noop.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3309
3310<h5>Example:</h5>
3311
3312<pre>
Dan Gohmanf54f50a2009-01-04 23:49:44 +0000[diff] [blame]3313 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3314 free [4 x i8]* %array
3315</pre>
3316</div>
3317
3318<!-- _______________________________________________________________________ -->
3319<div class="doc_subsubsection">
3320 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
3321</div>
3322
3323<div class="doc_text">
3324
3325<h5>Syntax:</h5>
3326
3327<pre>
3328 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
3329</pre>
3330
3331<h5>Overview:</h5>
3332
3333<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
3334currently executing function, to be automatically released when this function
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]3335returns to its caller. The object is always allocated in the generic address
3336space (address space zero).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3337
3338<h5>Arguments:</h5>
3339
3340<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
3341bytes of memory on the runtime stack, returning a pointer of the
Gabor Greif5082cf42008-02-09 22:24:34 +0000[diff] [blame]3342appropriate type to the program. If "NumElements" is specified, it is the
3343number of elements allocated, otherwise "NumElements" is defaulted to be one.
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3344If a constant alignment is specified, the value result of the allocation is guaranteed
Gabor Greif5082cf42008-02-09 22:24:34 +0000[diff] [blame]3345to be aligned to at least that boundary. If not specified, or if zero, the target
3346can choose to align the allocation on any convenient boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3347
3348<p>'<tt>type</tt>' may be any sized type.</p>
3349
3350<h5>Semantics:</h5>
3351
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]3352<p>Memory is allocated; a pointer is returned. The operation is undefiend if
3353there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3354memory is automatically released when the function returns. The '<tt>alloca</tt>'
3355instruction is commonly used to represent automatic variables that must
3356have an address available. When the function returns (either with the <tt><a
3357 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3358instructions), the memory is reclaimed. Allocating zero bytes
3359is legal, but the result is undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3360
3361<h5>Example:</h5>
3362
3363<pre>
Dan Gohmanf54f50a2009-01-04 23:49:44 +0000[diff] [blame]3364 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
3365 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
3366 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
3367 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3368</pre>
3369</div>
3370
3371<!-- _______________________________________________________________________ -->
3372<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
3373Instruction</a> </div>
3374<div class="doc_text">
3375<h5>Syntax:</h5>
3376<pre> &lt;result&gt; = load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br> &lt;result&gt; = volatile load &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;]<br></pre>
3377<h5>Overview:</h5>
3378<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
3379<h5>Arguments:</h5>
3380<p>The argument to the '<tt>load</tt>' instruction specifies the memory
3381address from which to load. The pointer must point to a <a
3382 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
3383marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
3384the number or order of execution of this <tt>load</tt> with other
3385volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
3386instructions. </p>
Chris Lattner21003c82008-01-06 21:04:43 +0000[diff] [blame]3387<p>
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3388The optional constant "align" argument specifies the alignment of the operation
Chris Lattner21003c82008-01-06 21:04:43 +0000[diff] [blame]3389(that is, the alignment of the memory address). A value of 0 or an
3390omitted "align" argument means that the operation has the preferential
3391alignment for the target. It is the responsibility of the code emitter
3392to ensure that the alignment information is correct. Overestimating
3393the alignment results in an undefined behavior. Underestimating the
3394alignment may produce less efficient code. An alignment of 1 is always
3395safe.
3396</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3397<h5>Semantics:</h5>
3398<p>The location of memory pointed to is loaded.</p>
3399<h5>Examples:</h5>
3400<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
3401 <a
3402 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
3403 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
3404</pre>
3405</div>
3406<!-- _______________________________________________________________________ -->
3407<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
3408Instruction</a> </div>
3409<div class="doc_text">
3410<h5>Syntax:</h5>
3411<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
3412 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
3413</pre>
3414<h5>Overview:</h5>
3415<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
3416<h5>Arguments:</h5>
3417<p>There are two arguments to the '<tt>store</tt>' instruction: a value
3418to store and an address at which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3419operand must be a pointer to the <a href="#t_firstclass">first class</a> type
3420of the '<tt>&lt;value&gt;</tt>'
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3421operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
3422optimizer is not allowed to modify the number or order of execution of
3423this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
3424 href="#i_store">store</a></tt> instructions.</p>
Chris Lattner21003c82008-01-06 21:04:43 +0000[diff] [blame]3425<p>
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3426The optional constant "align" argument specifies the alignment of the operation
Chris Lattner21003c82008-01-06 21:04:43 +0000[diff] [blame]3427(that is, the alignment of the memory address). A value of 0 or an
3428omitted "align" argument means that the operation has the preferential
3429alignment for the target. It is the responsibility of the code emitter
3430to ensure that the alignment information is correct. Overestimating
3431the alignment results in an undefined behavior. Underestimating the
3432alignment may produce less efficient code. An alignment of 1 is always
3433safe.
3434</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3435<h5>Semantics:</h5>
3436<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
3437at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
3438<h5>Example:</h5>
3439<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling63ffa142007-10-22 05:10:05 +0000[diff] [blame]3440 store i32 3, i32* %ptr <i>; yields {void}</i>
3441 %val = <a href="#i_load">load</a> i32* %ptr <i>; yields {i32}:val = i32 3</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3442</pre>
3443</div>
3444
3445<!-- _______________________________________________________________________ -->
3446<div class="doc_subsubsection">
3447 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
3448</div>
3449
3450<div class="doc_text">
3451<h5>Syntax:</h5>
3452<pre>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]3453 &lt;result&gt; = getelementptr &lt;pty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3454</pre>
3455
3456<h5>Overview:</h5>
3457
3458<p>
3459The '<tt>getelementptr</tt>' instruction is used to get the address of a
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]3460subelement of an aggregate data structure. It performs address calculation only
3461and does not access memory.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3462
3463<h5>Arguments:</h5>
3464
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]3465<p>The first argument is always a pointer, and forms the basis of the
3466calculation. The remaining arguments are indices, that indicate which of the
3467elements of the aggregate object are indexed. The interpretation of each index
3468is dependent on the type being indexed into. The first index always indexes the
3469pointer value given as the first argument, the second index indexes a value of
3470the type pointed to (not necessarily the value directly pointed to, since the
3471first index can be non-zero), etc. The first type indexed into must be a pointer
3472value, subsequent types can be arrays, vectors and structs. Note that subsequent
3473types being indexed into can never be pointers, since that would require loading
3474the pointer before continuing calculation.</p>
3475
3476<p>The type of each index argument depends on the type it is indexing into.
3477When indexing into a (packed) structure, only <tt>i32</tt> integer
3478<b>constants</b> are allowed. When indexing into an array, pointer or vector,
3479only integers of 32 or 64 bits are allowed (also non-constants). 32-bit values
3480will be sign extended to 64-bits if required.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3481
3482<p>For example, let's consider a C code fragment and how it gets
3483compiled to LLVM:</p>
3484
3485<div class="doc_code">
3486<pre>
3487struct RT {
3488 char A;
3489 int B[10][20];
3490 char C;
3491};
3492struct ST {
3493 int X;
3494 double Y;
3495 struct RT Z;
3496};
3497
3498int *foo(struct ST *s) {
3499 return &amp;s[1].Z.B[5][13];
3500}
3501</pre>
3502</div>
3503
3504<p>The LLVM code generated by the GCC frontend is:</p>
3505
3506<div class="doc_code">
3507<pre>
Chris Lattner5b6dc6e2009-01-11 20:53:49 +0000[diff] [blame]3508%RT = <a href="#namedtypes">type</a> { i8 , [10 x [20 x i32]], i8 }
3509%ST = <a href="#namedtypes">type</a> { i32, double, %RT }
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3510
3511define i32* %foo(%ST* %s) {
3512entry:
3513 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
3514 ret i32* %reg
3515}
3516</pre>
3517</div>
3518
3519<h5>Semantics:</h5>
3520
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3521<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
3522type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
3523}</tt>' type, a structure. The second index indexes into the third element of
3524the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
3525i8 }</tt>' type, another structure. The third index indexes into the second
3526element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
3527array. The two dimensions of the array are subscripted into, yielding an
3528'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
3529to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
3530
3531<p>Note that it is perfectly legal to index partially through a
3532structure, returning a pointer to an inner element. Because of this,
3533the LLVM code for the given testcase is equivalent to:</p>
3534
3535<pre>
3536 define i32* %foo(%ST* %s) {
3537 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
3538 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
3539 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
3540 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
3541 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
3542 ret i32* %t5
3543 }
3544</pre>
3545
3546<p>Note that it is undefined to access an array out of bounds: array and
3547pointer indexes must always be within the defined bounds of the array type.
Chris Lattnera7d94ba2008-04-24 05:59:56 +0000[diff] [blame]3548The one exception for this rule is zero length arrays. These arrays are
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3549defined to be accessible as variable length arrays, which requires access
3550beyond the zero'th element.</p>
3551
3552<p>The getelementptr instruction is often confusing. For some more insight
3553into how it works, see <a href="GetElementPtr.html">the getelementptr
3554FAQ</a>.</p>
3555
3556<h5>Example:</h5>
3557
3558<pre>
3559 <i>; yields [12 x i8]*:aptr</i>
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]3560 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
3561 <i>; yields i8*:vptr</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]3562 %vptr = getelementptr {i32, &lt;2 x i8&gt;}* %svptr, i64 0, i32 1, i32 1
Matthijs Kooijman32a080e2008-10-13 13:44:15 +0000[diff] [blame]3563 <i>; yields i8*:eptr</i>
3564 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3565</pre>
3566</div>
3567
3568<!-- ======================================================================= -->
3569<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
3570</div>
3571<div class="doc_text">
3572<p>The instructions in this category are the conversion instructions (casting)
3573which all take a single operand and a type. They perform various bit conversions
3574on the operand.</p>
3575</div>
3576
3577<!-- _______________________________________________________________________ -->
3578<div class="doc_subsubsection">
3579 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
3580</div>
3581<div class="doc_text">
3582
3583<h5>Syntax:</h5>
3584<pre>
3585 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3586</pre>
3587
3588<h5>Overview:</h5>
3589<p>
3590The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
3591</p>
3592
3593<h5>Arguments:</h5>
3594<p>
3595The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
3596be an <a href="#t_integer">integer</a> type, and a type that specifies the size
3597and type of the result, which must be an <a href="#t_integer">integer</a>
3598type. The bit size of <tt>value</tt> must be larger than the bit size of
3599<tt>ty2</tt>. Equal sized types are not allowed.</p>
3600
3601<h5>Semantics:</h5>
3602<p>
3603The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
3604and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
3605larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
3606It will always truncate bits.</p>
3607
3608<h5>Example:</h5>
3609<pre>
3610 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
3611 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
3612 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
3613</pre>
3614</div>
3615
3616<!-- _______________________________________________________________________ -->
3617<div class="doc_subsubsection">
3618 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
3619</div>
3620<div class="doc_text">
3621
3622<h5>Syntax:</h5>
3623<pre>
3624 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3625</pre>
3626
3627<h5>Overview:</h5>
3628<p>The '<tt>zext</tt>' instruction zero extends its operand to type
3629<tt>ty2</tt>.</p>
3630
3631
3632<h5>Arguments:</h5>
3633<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
3634<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3635also be of <a href="#t_integer">integer</a> type. The bit size of the
3636<tt>value</tt> must be smaller than the bit size of the destination type,
3637<tt>ty2</tt>.</p>
3638
3639<h5>Semantics:</h5>
3640<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
3641bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
3642
3643<p>When zero extending from i1, the result will always be either 0 or 1.</p>
3644
3645<h5>Example:</h5>
3646<pre>
3647 %X = zext i32 257 to i64 <i>; yields i64:257</i>
3648 %Y = zext i1 true to i32 <i>; yields i32:1</i>
3649</pre>
3650</div>
3651
3652<!-- _______________________________________________________________________ -->
3653<div class="doc_subsubsection">
3654 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
3655</div>
3656<div class="doc_text">
3657
3658<h5>Syntax:</h5>
3659<pre>
3660 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3661</pre>
3662
3663<h5>Overview:</h5>
3664<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
3665
3666<h5>Arguments:</h5>
3667<p>
3668The '<tt>sext</tt>' instruction takes a value to cast, which must be of
3669<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3670also be of <a href="#t_integer">integer</a> type. The bit size of the
3671<tt>value</tt> must be smaller than the bit size of the destination type,
3672<tt>ty2</tt>.</p>
3673
3674<h5>Semantics:</h5>
3675<p>
3676The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
3677bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
3678the type <tt>ty2</tt>.</p>
3679
3680<p>When sign extending from i1, the extension always results in -1 or 0.</p>
3681
3682<h5>Example:</h5>
3683<pre>
3684 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
3685 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
3686</pre>
3687</div>
3688
3689<!-- _______________________________________________________________________ -->
3690<div class="doc_subsubsection">
3691 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3692</div>
3693
3694<div class="doc_text">
3695
3696<h5>Syntax:</h5>
3697
3698<pre>
3699 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3700</pre>
3701
3702<h5>Overview:</h5>
3703<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
3704<tt>ty2</tt>.</p>
3705
3706
3707<h5>Arguments:</h5>
3708<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
3709 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
3710cast it to. The size of <tt>value</tt> must be larger than the size of
3711<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3712<i>no-op cast</i>.</p>
3713
3714<h5>Semantics:</h5>
3715<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3716<a href="#t_floating">floating point</a> type to a smaller
3717<a href="#t_floating">floating point</a> type. If the value cannot fit within
3718the destination type, <tt>ty2</tt>, then the results are undefined.</p>
3719
3720<h5>Example:</h5>
3721<pre>
3722 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3723 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3724</pre>
3725</div>
3726
3727<!-- _______________________________________________________________________ -->
3728<div class="doc_subsubsection">
3729 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
3730</div>
3731<div class="doc_text">
3732
3733<h5>Syntax:</h5>
3734<pre>
3735 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3736</pre>
3737
3738<h5>Overview:</h5>
3739<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
3740floating point value.</p>
3741
3742<h5>Arguments:</h5>
3743<p>The '<tt>fpext</tt>' instruction takes a
3744<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
3745and a <a href="#t_floating">floating point</a> type to cast it to. The source
3746type must be smaller than the destination type.</p>
3747
3748<h5>Semantics:</h5>
3749<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
3750<a href="#t_floating">floating point</a> type to a larger
3751<a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
3752used to make a <i>no-op cast</i> because it always changes bits. Use
3753<tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
3754
3755<h5>Example:</h5>
3756<pre>
3757 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
3758 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
3759</pre>
3760</div>
3761
3762<!-- _______________________________________________________________________ -->
3763<div class="doc_subsubsection">
3764 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
3765</div>
3766<div class="doc_text">
3767
3768<h5>Syntax:</h5>
3769<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3770 &lt;result&gt; = fptoui &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3771</pre>
3772
3773<h5>Overview:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3774<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3775unsigned integer equivalent of type <tt>ty2</tt>.
3776</p>
3777
3778<h5>Arguments:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3779<p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]3780scalar or vector <a href="#t_floating">floating point</a> value, and a type
3781to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3782type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3783vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3784
3785<h5>Semantics:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3786<p> The '<tt>fptoui</tt>' instruction converts its
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3787<a href="#t_floating">floating point</a> operand into the nearest (rounding
3788towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3789the results are undefined.</p>
3790
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3791<h5>Example:</h5>
3792<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3793 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]3794 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3795 %X = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3796</pre>
3797</div>
3798
3799<!-- _______________________________________________________________________ -->
3800<div class="doc_subsubsection">
3801 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
3802</div>
3803<div class="doc_text">
3804
3805<h5>Syntax:</h5>
3806<pre>
3807 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3808</pre>
3809
3810<h5>Overview:</h5>
3811<p>The '<tt>fptosi</tt>' instruction converts
3812<a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
3813</p>
3814
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3815<h5>Arguments:</h5>
3816<p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]3817scalar or vector <a href="#t_floating">floating point</a> value, and a type
3818to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3819type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3820vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3821
3822<h5>Semantics:</h5>
3823<p>The '<tt>fptosi</tt>' instruction converts its
3824<a href="#t_floating">floating point</a> operand into the nearest (rounding
3825towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3826the results are undefined.</p>
3827
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3828<h5>Example:</h5>
3829<pre>
3830 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]3831 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3832 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
3833</pre>
3834</div>
3835
3836<!-- _______________________________________________________________________ -->
3837<div class="doc_subsubsection">
3838 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
3839</div>
3840<div class="doc_text">
3841
3842<h5>Syntax:</h5>
3843<pre>
3844 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3845</pre>
3846
3847<h5>Overview:</h5>
3848<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
3849integer and converts that value to the <tt>ty2</tt> type.</p>
3850
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3851<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]3852<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a
3853scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3854to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3855type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3856floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3857
3858<h5>Semantics:</h5>
3859<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
3860integer quantity and converts it to the corresponding floating point value. If
3861the value cannot fit in the floating point value, the results are undefined.</p>
3862
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3863<h5>Example:</h5>
3864<pre>
3865 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]3866 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3867</pre>
3868</div>
3869
3870<!-- _______________________________________________________________________ -->
3871<div class="doc_subsubsection">
3872 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
3873</div>
3874<div class="doc_text">
3875
3876<h5>Syntax:</h5>
3877<pre>
3878 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3879</pre>
3880
3881<h5>Overview:</h5>
3882<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
3883integer and converts that value to the <tt>ty2</tt> type.</p>
3884
3885<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]3886<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a
3887scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3888to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3889type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3890floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3891
3892<h5>Semantics:</h5>
3893<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
3894integer quantity and converts it to the corresponding floating point value. If
3895the value cannot fit in the floating point value, the results are undefined.</p>
3896
3897<h5>Example:</h5>
3898<pre>
3899 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]3900 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3901</pre>
3902</div>
3903
3904<!-- _______________________________________________________________________ -->
3905<div class="doc_subsubsection">
3906 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3907</div>
3908<div class="doc_text">
3909
3910<h5>Syntax:</h5>
3911<pre>
3912 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3913</pre>
3914
3915<h5>Overview:</h5>
3916<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3917the integer type <tt>ty2</tt>.</p>
3918
3919<h5>Arguments:</h5>
3920<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
3921must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]3922<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3923
3924<h5>Semantics:</h5>
3925<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3926<tt>ty2</tt> by interpreting the pointer value as an integer and either
3927truncating or zero extending that value to the size of the integer type. If
3928<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3929<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
3930are the same size, then nothing is done (<i>no-op cast</i>) other than a type
3931change.</p>
3932
3933<h5>Example:</h5>
3934<pre>
3935 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
3936 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
3937</pre>
3938</div>
3939
3940<!-- _______________________________________________________________________ -->
3941<div class="doc_subsubsection">
3942 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3943</div>
3944<div class="doc_text">
3945
3946<h5>Syntax:</h5>
3947<pre>
3948 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3949</pre>
3950
3951<h5>Overview:</h5>
3952<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3953a pointer type, <tt>ty2</tt>.</p>
3954
3955<h5>Arguments:</h5>
3956<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
3957value to cast, and a type to cast it to, which must be a
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]3958<a href="#t_pointer">pointer</a> type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3959
3960<h5>Semantics:</h5>
3961<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3962<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3963the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3964size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3965the size of a pointer then a zero extension is done. If they are the same size,
3966nothing is done (<i>no-op cast</i>).</p>
3967
3968<h5>Example:</h5>
3969<pre>
3970 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
3971 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
3972 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
3973</pre>
3974</div>
3975
3976<!-- _______________________________________________________________________ -->
3977<div class="doc_subsubsection">
3978 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
3979</div>
3980<div class="doc_text">
3981
3982<h5>Syntax:</h5>
3983<pre>
3984 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3985</pre>
3986
3987<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3988
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3989<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
3990<tt>ty2</tt> without changing any bits.</p>
3991
3992<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3993
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3994<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
Dan Gohman7305fa02008-09-08 16:45:59 +0000[diff] [blame]3995a non-aggregate first class value, and a type to cast it to, which must also be
3996a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes of
3997<tt>value</tt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3998and the destination type, <tt>ty2</tt>, must be identical. If the source
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3999type is a pointer, the destination type must also be a pointer. This
4000instruction supports bitwise conversion of vectors to integers and to vectors
4001of other types (as long as they have the same size).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4002
4003<h5>Semantics:</h5>
4004<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
4005<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
4006this conversion. The conversion is done as if the <tt>value</tt> had been
4007stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
4008converted to other pointer types with this instruction. To convert pointers to
4009other types, use the <a href="#i_inttoptr">inttoptr</a> or
4010<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
4011
4012<h5>Example:</h5>
4013<pre>
4014 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
4015 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4016 %Z = bitcast &lt;2 x int&gt; %V to i64; <i>; yields i64: %V</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4017</pre>
4018</div>
4019
4020<!-- ======================================================================= -->
4021<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
4022<div class="doc_text">
4023<p>The instructions in this category are the "miscellaneous"
4024instructions, which defy better classification.</p>
4025</div>
4026
4027<!-- _______________________________________________________________________ -->
4028<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
4029</div>
4030<div class="doc_text">
4031<h5>Syntax:</h5>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4032<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {i1} or {&lt;N x i1&gt;}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4033</pre>
4034<h5>Overview:</h5>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4035<p>The '<tt>icmp</tt>' instruction returns a boolean value or
4036a vector of boolean values based on comparison
4037of its two integer, integer vector, or pointer operands.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4038<h5>Arguments:</h5>
4039<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
4040the condition code indicating the kind of comparison to perform. It is not
4041a value, just a keyword. The possible condition code are:
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4042</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4043<ol>
4044 <li><tt>eq</tt>: equal</li>
4045 <li><tt>ne</tt>: not equal </li>
4046 <li><tt>ugt</tt>: unsigned greater than</li>
4047 <li><tt>uge</tt>: unsigned greater or equal</li>
4048 <li><tt>ult</tt>: unsigned less than</li>
4049 <li><tt>ule</tt>: unsigned less or equal</li>
4050 <li><tt>sgt</tt>: signed greater than</li>
4051 <li><tt>sge</tt>: signed greater or equal</li>
4052 <li><tt>slt</tt>: signed less than</li>
4053 <li><tt>sle</tt>: signed less or equal</li>
4054</ol>
4055<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4056<a href="#t_pointer">pointer</a>
4057or integer <a href="#t_vector">vector</a> typed.
4058They must also be identical types.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4059<h5>Semantics:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4060<p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4061the condition code given as <tt>cond</tt>. The comparison performed always
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4062yields either an <a href="#t_primitive"><tt>i1</tt></a> or vector of <tt>i1</tt> result, as follows:
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4063</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4064<ol>
4065 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
4066 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
4067 </li>
4068 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4069 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4070 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4071 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4072 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4073 <tt>true</tt> if <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4074 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4075 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4076 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4077 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4078 <li><tt>sgt</tt>: interprets the operands as signed values and yields
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4079 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4080 <li><tt>sge</tt>: interprets the operands as signed values and yields
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4081 <tt>true</tt> if <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4082 <li><tt>slt</tt>: interprets the operands as signed values and yields
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4083 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4084 <li><tt>sle</tt>: interprets the operands as signed values and yields
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4085 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4086</ol>
4087<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
4088values are compared as if they were integers.</p>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4089<p>If the operands are integer vectors, then they are compared
4090element by element. The result is an <tt>i1</tt> vector with
4091the same number of elements as the values being compared.
4092Otherwise, the result is an <tt>i1</tt>.
4093</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4094
4095<h5>Example:</h5>
4096<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
4097 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
4098 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
4099 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
4100 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
4101 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
4102</pre>
Dan Gohmana53eb382009-01-22 01:39:38 +0000[diff] [blame]4103
4104<p>Note that the code generator does not yet support vector types with
4105 the <tt>icmp</tt> instruction.</p>
4106
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4107</div>
4108
4109<!-- _______________________________________________________________________ -->
4110<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
4111</div>
4112<div class="doc_text">
4113<h5>Syntax:</h5>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4114<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {i1} or {&lt;N x i1&gt;}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4115</pre>
4116<h5>Overview:</h5>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4117<p>The '<tt>fcmp</tt>' instruction returns a boolean value
4118or vector of boolean values based on comparison
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4119of its operands.</p>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4120<p>
4121If the operands are floating point scalars, then the result
4122type is a boolean (<a href="#t_primitive"><tt>i1</tt></a>).
4123</p>
4124<p>If the operands are floating point vectors, then the result type
4125is a vector of boolean with the same number of elements as the
4126operands being compared.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4127<h5>Arguments:</h5>
4128<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
4129the condition code indicating the kind of comparison to perform. It is not
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4130a value, just a keyword. The possible condition code are:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4131<ol>
4132 <li><tt>false</tt>: no comparison, always returns false</li>
4133 <li><tt>oeq</tt>: ordered and equal</li>
4134 <li><tt>ogt</tt>: ordered and greater than </li>
4135 <li><tt>oge</tt>: ordered and greater than or equal</li>
4136 <li><tt>olt</tt>: ordered and less than </li>
4137 <li><tt>ole</tt>: ordered and less than or equal</li>
4138 <li><tt>one</tt>: ordered and not equal</li>
4139 <li><tt>ord</tt>: ordered (no nans)</li>
4140 <li><tt>ueq</tt>: unordered or equal</li>
4141 <li><tt>ugt</tt>: unordered or greater than </li>
4142 <li><tt>uge</tt>: unordered or greater than or equal</li>
4143 <li><tt>ult</tt>: unordered or less than </li>
4144 <li><tt>ule</tt>: unordered or less than or equal</li>
4145 <li><tt>une</tt>: unordered or not equal</li>
4146 <li><tt>uno</tt>: unordered (either nans)</li>
4147 <li><tt>true</tt>: no comparison, always returns true</li>
4148</ol>
4149<p><i>Ordered</i> means that neither operand is a QNAN while
4150<i>unordered</i> means that either operand may be a QNAN.</p>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4151<p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be
4152either a <a href="#t_floating">floating point</a> type
4153or a <a href="#t_vector">vector</a> of floating point type.
4154They must have identical types.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4155<h5>Semantics:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4156<p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4157according to the condition code given as <tt>cond</tt>.
4158If the operands are vectors, then the vectors are compared
4159element by element.
4160Each comparison performed
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4161always yields an <a href="#t_primitive">i1</a> result, as follows:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4162<ol>
4163 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
4164 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4165 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4166 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4167 <tt>op1</tt> is greather than <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4168 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4169 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4170 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4171 <tt>op1</tt> is less than <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4172 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4173 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4174 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4175 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4176 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
4177 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4178 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4179 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4180 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4181 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4182 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4183 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4184 <tt>op1</tt> is less than <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4185 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4186 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4187 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4188 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4189 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
4190 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
4191</ol>
4192
4193<h5>Example:</h5>
4194<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4195 &lt;result&gt; = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
4196 &lt;result&gt; = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
4197 &lt;result&gt; = fcmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4198</pre>
Dan Gohmana53eb382009-01-22 01:39:38 +0000[diff] [blame]4199
4200<p>Note that the code generator does not yet support vector types with
4201 the <tt>fcmp</tt> instruction.</p>
4202
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4203</div>
4204
4205<!-- _______________________________________________________________________ -->
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4206<div class="doc_subsubsection">
4207 <a name="i_vicmp">'<tt>vicmp</tt>' Instruction</a>
4208</div>
4209<div class="doc_text">
4210<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4211<pre> &lt;result&gt; = vicmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt; <i>; yields {ty}:result</i>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4212</pre>
4213<h5>Overview:</h5>
4214<p>The '<tt>vicmp</tt>' instruction returns an integer vector value based on
4215element-wise comparison of its two integer vector operands.</p>
4216<h5>Arguments:</h5>
4217<p>The '<tt>vicmp</tt>' instruction takes three operands. The first operand is
4218the condition code indicating the kind of comparison to perform. It is not
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4219a value, just a keyword. The possible condition code are:</p>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4220<ol>
4221 <li><tt>eq</tt>: equal</li>
4222 <li><tt>ne</tt>: not equal </li>
4223 <li><tt>ugt</tt>: unsigned greater than</li>
4224 <li><tt>uge</tt>: unsigned greater or equal</li>
4225 <li><tt>ult</tt>: unsigned less than</li>
4226 <li><tt>ule</tt>: unsigned less or equal</li>
4227 <li><tt>sgt</tt>: signed greater than</li>
4228 <li><tt>sge</tt>: signed greater or equal</li>
4229 <li><tt>slt</tt>: signed less than</li>
4230 <li><tt>sle</tt>: signed less or equal</li>
4231</ol>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4232<p>The remaining two arguments must be <a href="#t_vector">vector</a> or
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4233<a href="#t_integer">integer</a> typed. They must also be identical types.</p>
4234<h5>Semantics:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4235<p>The '<tt>vicmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4236according to the condition code given as <tt>cond</tt>. The comparison yields a
4237<a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, of
4238identical type as the values being compared. The most significant bit in each
4239element is 1 if the element-wise comparison evaluates to true, and is 0
4240otherwise. All other bits of the result are undefined. The condition codes
4241are evaluated identically to the <a href="#i_icmp">'<tt>icmp</tt>'
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4242instruction</a>.</p>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4243
4244<h5>Example:</h5>
4245<pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4246 &lt;result&gt; = vicmp eq &lt;2 x i32&gt; &lt; i32 4, i32 0&gt;, &lt; i32 5, i32 0&gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0, i32 -1 &gt;</i>
4247 &lt;result&gt; = vicmp ult &lt;2 x i8 &gt; &lt; i8 1, i8 2&gt;, &lt; i8 2, i8 2 &gt; <i>; yields: result=&lt;2 x i8&gt; &lt; i8 -1, i8 0 &gt;</i>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4248</pre>
4249</div>
4250
4251<!-- _______________________________________________________________________ -->
4252<div class="doc_subsubsection">
4253 <a name="i_vfcmp">'<tt>vfcmp</tt>' Instruction</a>
4254</div>
4255<div class="doc_text">
4256<h5>Syntax:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4257<pre> &lt;result&gt; = vfcmp &lt;cond&gt; &lt;ty&gt; &lt;op1&gt;, &lt;op2&gt;</pre>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4258<h5>Overview:</h5>
4259<p>The '<tt>vfcmp</tt>' instruction returns an integer vector value based on
4260element-wise comparison of its two floating point vector operands. The output
4261elements have the same width as the input elements.</p>
4262<h5>Arguments:</h5>
4263<p>The '<tt>vfcmp</tt>' instruction takes three operands. The first operand is
4264the condition code indicating the kind of comparison to perform. It is not
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4265a value, just a keyword. The possible condition code are:</p>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4266<ol>
4267 <li><tt>false</tt>: no comparison, always returns false</li>
4268 <li><tt>oeq</tt>: ordered and equal</li>
4269 <li><tt>ogt</tt>: ordered and greater than </li>
4270 <li><tt>oge</tt>: ordered and greater than or equal</li>
4271 <li><tt>olt</tt>: ordered and less than </li>
4272 <li><tt>ole</tt>: ordered and less than or equal</li>
4273 <li><tt>one</tt>: ordered and not equal</li>
4274 <li><tt>ord</tt>: ordered (no nans)</li>
4275 <li><tt>ueq</tt>: unordered or equal</li>
4276 <li><tt>ugt</tt>: unordered or greater than </li>
4277 <li><tt>uge</tt>: unordered or greater than or equal</li>
4278 <li><tt>ult</tt>: unordered or less than </li>
4279 <li><tt>ule</tt>: unordered or less than or equal</li>
4280 <li><tt>une</tt>: unordered or not equal</li>
4281 <li><tt>uno</tt>: unordered (either nans)</li>
4282 <li><tt>true</tt>: no comparison, always returns true</li>
4283</ol>
4284<p>The remaining two arguments must be <a href="#t_vector">vector</a> of
4285<a href="#t_floating">floating point</a> typed. They must also be identical
4286types.</p>
4287<h5>Semantics:</h5>
Gabor Greifd9068fe2008-08-07 21:46:00 +0000[diff] [blame]4288<p>The '<tt>vfcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4289according to the condition code given as <tt>cond</tt>. The comparison yields a
4290<a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, with
4291an identical number of elements as the values being compared, and each element
4292having identical with to the width of the floating point elements. The most
4293significant bit in each element is 1 if the element-wise comparison evaluates to
4294true, and is 0 otherwise. All other bits of the result are undefined. The
4295condition codes are evaluated identically to the
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4296<a href="#i_fcmp">'<tt>fcmp</tt>' instruction</a>.</p>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4297
4298<h5>Example:</h5>
4299<pre>
Chris Lattner1e5c5cd02008-10-13 16:55:18 +0000[diff] [blame]4300 <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0, i32 -1 &gt;</i>
4301 &lt;result&gt; = vfcmp oeq &lt;2 x float&gt; &lt; float 4, float 0 &gt;, &lt; float 5, float 0 &gt;
4302
4303 <i>; yields: result=&lt;2 x i64&gt; &lt; i64 -1, i64 0 &gt;</i>
4304 &lt;result&gt; = vfcmp ult &lt;2 x double&gt; &lt; double 1, double 2 &gt;, &lt; double 2, double 2&gt;
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4305</pre>
4306</div>
4307
4308<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4309<div class="doc_subsubsection">
4310 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
4311</div>
4312
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4313<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4314
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4315<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4316
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4317<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
4318<h5>Overview:</h5>
4319<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
4320the SSA graph representing the function.</p>
4321<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4322
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4323<p>The type of the incoming values is specified with the first type
4324field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
4325as arguments, with one pair for each predecessor basic block of the
4326current block. Only values of <a href="#t_firstclass">first class</a>
4327type may be used as the value arguments to the PHI node. Only labels
4328may be used as the label arguments.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4329
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4330<p>There must be no non-phi instructions between the start of a basic
4331block and the PHI instructions: i.e. PHI instructions must be first in
4332a basic block.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4333
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4334<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4335
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4336<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
4337specified by the pair corresponding to the predecessor basic block that executed
4338just prior to the current block.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4339
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4340<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4341<pre>
4342Loop: ; Infinite loop that counts from 0 on up...
4343 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
4344 %nextindvar = add i32 %indvar, 1
4345 br label %Loop
4346</pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4347</div>
4348
4349<!-- _______________________________________________________________________ -->
4350<div class="doc_subsubsection">
4351 <a name="i_select">'<tt>select</tt>' Instruction</a>
4352</div>
4353
4354<div class="doc_text">
4355
4356<h5>Syntax:</h5>
4357
4358<pre>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4359 &lt;result&gt; = select <i>selty</i> &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
4360
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4361 <i>selty</i> is either i1 or {&lt;N x i1&gt;}
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4362</pre>
4363
4364<h5>Overview:</h5>
4365
4366<p>
4367The '<tt>select</tt>' instruction is used to choose one value based on a
4368condition, without branching.
4369</p>
4370
4371
4372<h5>Arguments:</h5>
4373
4374<p>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4375The '<tt>select</tt>' instruction requires an 'i1' value or
4376a vector of 'i1' values indicating the
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4377condition, and two values of the same <a href="#t_firstclass">first class</a>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4378type. If the val1/val2 are vectors and
4379the condition is a scalar, then entire vectors are selected, not
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4380individual elements.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4381</p>
4382
4383<h5>Semantics:</h5>
4384
4385<p>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4386If the condition is an i1 and it evaluates to 1, the instruction returns the first
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4387value argument; otherwise, it returns the second value argument.
4388</p>
Dan Gohmanb60ca3c2008-09-09 01:02:47 +0000[diff] [blame]4389<p>
4390If the condition is a vector of i1, then the value arguments must
4391be vectors of the same size, and the selection is done element
4392by element.
4393</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4394
4395<h5>Example:</h5>
4396
4397<pre>
4398 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
4399</pre>
Dan Gohmana53eb382009-01-22 01:39:38 +0000[diff] [blame]4400
4401<p>Note that the code generator does not yet support conditions
4402 with vector type.</p>
4403
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4404</div>
4405
4406
4407<!-- _______________________________________________________________________ -->
4408<div class="doc_subsubsection">
4409 <a name="i_call">'<tt>call</tt>' Instruction</a>
4410</div>
4411
4412<div class="doc_text">
4413
4414<h5>Syntax:</h5>
4415<pre>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]4416 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] &lt;ty&gt; [&lt;fnty&gt;*] &lt;fnptrval&gt;(&lt;function args&gt;) [<a href="#fnattrs">fn attrs</a>]
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4417</pre>
4418
4419<h5>Overview:</h5>
4420
4421<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
4422
4423<h5>Arguments:</h5>
4424
4425<p>This instruction requires several arguments:</p>
4426
4427<ol>
4428 <li>
4429 <p>The optional "tail" marker indicates whether the callee function accesses
4430 any allocas or varargs in the caller. If the "tail" marker is present, the
4431 function call is eligible for tail call optimization. Note that calls may
4432 be marked "tail" even if they do not occur before a <a
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4433 href="#i_ret"><tt>ret</tt></a> instruction.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4434 </li>
4435 <li>
4436 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
4437 convention</a> the call should use. If none is specified, the call defaults
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4438 to using C calling conventions.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4439 </li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]4440
4441 <li>
4442 <p>The optional <a href="#paramattrs">Parameter Attributes</a> list for
4443 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>',
4444 and '<tt>inreg</tt>' attributes are valid here.</p>
4445 </li>
4446
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4447 <li>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]4448 <p>'<tt>ty</tt>': the type of the call instruction itself which is also
4449 the type of the return value. Functions that return no value are marked
4450 <tt><a href="#t_void">void</a></tt>.</p>
4451 </li>
4452 <li>
4453 <p>'<tt>fnty</tt>': shall be the signature of the pointer to function
4454 value being invoked. The argument types must match the types implied by
4455 this signature. This type can be omitted if the function is not varargs
4456 and if the function type does not return a pointer to a function.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4457 </li>
4458 <li>
4459 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
4460 be invoked. In most cases, this is a direct function invocation, but
4461 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
4462 to function value.</p>
4463 </li>
4464 <li>
4465 <p>'<tt>function args</tt>': argument list whose types match the
4466 function signature argument types. All arguments must be of
4467 <a href="#t_firstclass">first class</a> type. If the function signature
4468 indicates the function accepts a variable number of arguments, the extra
4469 arguments can be specified.</p>
4470 </li>
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]4471 <li>
Devang Pateld0bfcc72008-10-07 17:48:33 +0000[diff] [blame]4472 <p>The optional <a href="#fnattrs">function attributes</a> list. Only
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]4473 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
4474 '<tt>readnone</tt>' attributes are valid here.</p>
4475 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4476</ol>
4477
4478<h5>Semantics:</h5>
4479
4480<p>The '<tt>call</tt>' instruction is used to cause control flow to
4481transfer to a specified function, with its incoming arguments bound to
4482the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
4483instruction in the called function, control flow continues with the
4484instruction after the function call, and the return value of the
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4485function is bound to the result argument.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4486
4487<h5>Example:</h5>
4488
4489<pre>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]4490 %retval = call i32 @test(i32 %argc)
Chris Lattner5e893ef2008-03-21 17:24:17 +0000[diff] [blame]4491 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42) <i>; yields i32</i>
4492 %X = tail call i32 @foo() <i>; yields i32</i>
4493 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
4494 call void %foo(i8 97 signext)
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4495
4496 %struct.A = type { i32, i8 }
Devang Patelac2fc272008-10-06 18:50:38 +0000[diff] [blame]4497 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
Dan Gohman3e700032008-10-04 19:00:07 +0000[diff] [blame]4498 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
4499 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
Chris Lattnerac454b32008-10-08 06:26:11 +0000[diff] [blame]4500 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
Matthijs Kooijman2c4e05a2008-10-07 10:03:45 +0000[diff] [blame]4501 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4502</pre>
4503
4504</div>
4505
4506<!-- _______________________________________________________________________ -->
4507<div class="doc_subsubsection">
4508 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
4509</div>
4510
4511<div class="doc_text">
4512
4513<h5>Syntax:</h5>
4514
4515<pre>
4516 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
4517</pre>
4518
4519<h5>Overview:</h5>
4520
4521<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
4522the "variable argument" area of a function call. It is used to implement the
4523<tt>va_arg</tt> macro in C.</p>
4524
4525<h5>Arguments:</h5>
4526
4527<p>This instruction takes a <tt>va_list*</tt> value and the type of
4528the argument. It returns a value of the specified argument type and
4529increments the <tt>va_list</tt> to point to the next argument. The
4530actual type of <tt>va_list</tt> is target specific.</p>
4531
4532<h5>Semantics:</h5>
4533
4534<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
4535type from the specified <tt>va_list</tt> and causes the
4536<tt>va_list</tt> to point to the next argument. For more information,
4537see the variable argument handling <a href="#int_varargs">Intrinsic
4538Functions</a>.</p>
4539
4540<p>It is legal for this instruction to be called in a function which does not
4541take a variable number of arguments, for example, the <tt>vfprintf</tt>
4542function.</p>
4543
4544<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
4545href="#intrinsics">intrinsic function</a> because it takes a type as an
4546argument.</p>
4547
4548<h5>Example:</h5>
4549
4550<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
4551
Dan Gohman60967192009-01-12 23:12:39 +0000[diff] [blame]4552<p>Note that the code generator does not yet fully support va_arg
4553 on many targets. Also, it does not currently support va_arg with
4554 aggregate types on any target.</p>
4555
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4556</div>
4557
4558<!-- *********************************************************************** -->
4559<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
4560<!-- *********************************************************************** -->
4561
4562<div class="doc_text">
4563
4564<p>LLVM supports the notion of an "intrinsic function". These functions have
4565well known names and semantics and are required to follow certain restrictions.
4566Overall, these intrinsics represent an extension mechanism for the LLVM
4567language that does not require changing all of the transformations in LLVM when
4568adding to the language (or the bitcode reader/writer, the parser, etc...).</p>
4569
4570<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
4571prefix is reserved in LLVM for intrinsic names; thus, function names may not
4572begin with this prefix. Intrinsic functions must always be external functions:
4573you cannot define the body of intrinsic functions. Intrinsic functions may
4574only be used in call or invoke instructions: it is illegal to take the address
4575of an intrinsic function. Additionally, because intrinsic functions are part
4576of the LLVM language, it is required if any are added that they be documented
4577here.</p>
4578
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4579<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
4580a family of functions that perform the same operation but on different data
4581types. Because LLVM can represent over 8 million different integer types,
4582overloading is used commonly to allow an intrinsic function to operate on any
4583integer type. One or more of the argument types or the result type can be
4584overloaded to accept any integer type. Argument types may also be defined as
4585exactly matching a previous argument's type or the result type. This allows an
4586intrinsic function which accepts multiple arguments, but needs all of them to
4587be of the same type, to only be overloaded with respect to a single argument or
4588the result.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4589
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4590<p>Overloaded intrinsics will have the names of its overloaded argument types
4591encoded into its function name, each preceded by a period. Only those types
4592which are overloaded result in a name suffix. Arguments whose type is matched
4593against another type do not. For example, the <tt>llvm.ctpop</tt> function can
4594take an integer of any width and returns an integer of exactly the same integer
4595width. This leads to a family of functions such as
4596<tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29 %val)</tt>.
4597Only one type, the return type, is overloaded, and only one type suffix is
4598required. Because the argument's type is matched against the return type, it
4599does not require its own name suffix.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4600
4601<p>To learn how to add an intrinsic function, please see the
4602<a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
4603</p>
4604
4605</div>
4606
4607<!-- ======================================================================= -->
4608<div class="doc_subsection">
4609 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
4610</div>
4611
4612<div class="doc_text">
4613
4614<p>Variable argument support is defined in LLVM with the <a
4615 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
4616intrinsic functions. These functions are related to the similarly
4617named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
4618
4619<p>All of these functions operate on arguments that use a
4620target-specific value type "<tt>va_list</tt>". The LLVM assembly
4621language reference manual does not define what this type is, so all
4622transformations should be prepared to handle these functions regardless of
4623the type used.</p>
4624
4625<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
4626instruction and the variable argument handling intrinsic functions are
4627used.</p>
4628
4629<div class="doc_code">
4630<pre>
4631define i32 @test(i32 %X, ...) {
4632 ; Initialize variable argument processing
4633 %ap = alloca i8*
4634 %ap2 = bitcast i8** %ap to i8*
4635 call void @llvm.va_start(i8* %ap2)
4636
4637 ; Read a single integer argument
4638 %tmp = va_arg i8** %ap, i32
4639
4640 ; Demonstrate usage of llvm.va_copy and llvm.va_end
4641 %aq = alloca i8*
4642 %aq2 = bitcast i8** %aq to i8*
4643 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
4644 call void @llvm.va_end(i8* %aq2)
4645
4646 ; Stop processing of arguments.
4647 call void @llvm.va_end(i8* %ap2)
4648 ret i32 %tmp
4649}
4650
4651declare void @llvm.va_start(i8*)
4652declare void @llvm.va_copy(i8*, i8*)
4653declare void @llvm.va_end(i8*)
4654</pre>
4655</div>
4656
4657</div>
4658
4659<!-- _______________________________________________________________________ -->
4660<div class="doc_subsubsection">
4661 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
4662</div>
4663
4664
4665<div class="doc_text">
4666<h5>Syntax:</h5>
4667<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
4668<h5>Overview:</h5>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4669<p>The '<tt>llvm.va_start</tt>' intrinsic initializes
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4670<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
4671href="#i_va_arg">va_arg</a></tt>.</p>
4672
4673<h5>Arguments:</h5>
4674
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4675<p>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4676
4677<h5>Semantics:</h5>
4678
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]4679<p>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4680macro available in C. In a target-dependent way, it initializes the
4681<tt>va_list</tt> element to which the argument points, so that the next call to
4682<tt>va_arg</tt> will produce the first variable argument passed to the function.
4683Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
4684last argument of the function as the compiler can figure that out.</p>
4685
4686</div>
4687
4688<!-- _______________________________________________________________________ -->
4689<div class="doc_subsubsection">
4690 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
4691</div>
4692
4693<div class="doc_text">
4694<h5>Syntax:</h5>
4695<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
4696<h5>Overview:</h5>
4697
4698<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
4699which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
4700or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
4701
4702<h5>Arguments:</h5>
4703
4704<p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
4705
4706<h5>Semantics:</h5>
4707
4708<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
4709macro available in C. In a target-dependent way, it destroys the
4710<tt>va_list</tt> element to which the argument points. Calls to <a
4711href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
4712<tt>llvm.va_copy</tt></a> must be matched exactly with calls to
4713<tt>llvm.va_end</tt>.</p>
4714
4715</div>
4716
4717<!-- _______________________________________________________________________ -->
4718<div class="doc_subsubsection">
4719 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
4720</div>
4721
4722<div class="doc_text">
4723
4724<h5>Syntax:</h5>
4725
4726<pre>
4727 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
4728</pre>
4729
4730<h5>Overview:</h5>
4731
4732<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
4733from the source argument list to the destination argument list.</p>
4734
4735<h5>Arguments:</h5>
4736
4737<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
4738The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
4739
4740
4741<h5>Semantics:</h5>
4742
4743<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
4744macro available in C. In a target-dependent way, it copies the source
4745<tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
4746intrinsic is necessary because the <tt><a href="#int_va_start">
4747llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
4748example, memory allocation.</p>
4749
4750</div>
4751
4752<!-- ======================================================================= -->
4753<div class="doc_subsection">
4754 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
4755</div>
4756
4757<div class="doc_text">
4758
4759<p>
4760LLVM support for <a href="GarbageCollection.html">Accurate Garbage
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]4761Collection</a> (GC) requires the implementation and generation of these
4762intrinsics.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4763These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
4764stack</a>, as well as garbage collector implementations that require <a
4765href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
4766Front-ends for type-safe garbage collected languages should generate these
4767intrinsics to make use of the LLVM garbage collectors. For more details, see <a
4768href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
4769</p>
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]4770
4771<p>The garbage collection intrinsics only operate on objects in the generic
4772 address space (address space zero).</p>
4773
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4774</div>
4775
4776<!-- _______________________________________________________________________ -->
4777<div class="doc_subsubsection">
4778 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
4779</div>
4780
4781<div class="doc_text">
4782
4783<h5>Syntax:</h5>
4784
4785<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4786 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4787</pre>
4788
4789<h5>Overview:</h5>
4790
4791<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
4792the code generator, and allows some metadata to be associated with it.</p>
4793
4794<h5>Arguments:</h5>
4795
4796<p>The first argument specifies the address of a stack object that contains the
4797root pointer. The second pointer (which must be either a constant or a global
4798value address) contains the meta-data to be associated with the root.</p>
4799
4800<h5>Semantics:</h5>
4801
Chris Lattnera7d94ba2008-04-24 05:59:56 +0000[diff] [blame]4802<p>At runtime, a call to this intrinsic stores a null pointer into the "ptrloc"
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4803location. At compile-time, the code generator generates information to allow
Gordon Henriksen40393542007-12-25 02:31:26 +0000[diff] [blame]4804the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
4805intrinsic may only be used in a function which <a href="#gc">specifies a GC
4806algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4807
4808</div>
4809
4810
4811<!-- _______________________________________________________________________ -->
4812<div class="doc_subsubsection">
4813 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
4814</div>
4815
4816<div class="doc_text">
4817
4818<h5>Syntax:</h5>
4819
4820<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4821 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4822</pre>
4823
4824<h5>Overview:</h5>
4825
4826<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
4827locations, allowing garbage collector implementations that require read
4828barriers.</p>
4829
4830<h5>Arguments:</h5>
4831
4832<p>The second argument is the address to read from, which should be an address
4833allocated from the garbage collector. The first object is a pointer to the
4834start of the referenced object, if needed by the language runtime (otherwise
4835null).</p>
4836
4837<h5>Semantics:</h5>
4838
4839<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
4840instruction, but may be replaced with substantially more complex code by the
Gordon Henriksen40393542007-12-25 02:31:26 +0000[diff] [blame]4841garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
4842may only be used in a function which <a href="#gc">specifies a GC
4843algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4844
4845</div>
4846
4847
4848<!-- _______________________________________________________________________ -->
4849<div class="doc_subsubsection">
4850 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
4851</div>
4852
4853<div class="doc_text">
4854
4855<h5>Syntax:</h5>
4856
4857<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4858 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4859</pre>
4860
4861<h5>Overview:</h5>
4862
4863<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
4864locations, allowing garbage collector implementations that require write
4865barriers (such as generational or reference counting collectors).</p>
4866
4867<h5>Arguments:</h5>
4868
4869<p>The first argument is the reference to store, the second is the start of the
4870object to store it to, and the third is the address of the field of Obj to
4871store to. If the runtime does not require a pointer to the object, Obj may be
4872null.</p>
4873
4874<h5>Semantics:</h5>
4875
4876<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
4877instruction, but may be replaced with substantially more complex code by the
Gordon Henriksen40393542007-12-25 02:31:26 +0000[diff] [blame]4878garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
4879may only be used in a function which <a href="#gc">specifies a GC
4880algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4881
4882</div>
4883
4884
4885
4886<!-- ======================================================================= -->
4887<div class="doc_subsection">
4888 <a name="int_codegen">Code Generator Intrinsics</a>
4889</div>
4890
4891<div class="doc_text">
4892<p>
4893These intrinsics are provided by LLVM to expose special features that may only
4894be implemented with code generator support.
4895</p>
4896
4897</div>
4898
4899<!-- _______________________________________________________________________ -->
4900<div class="doc_subsubsection">
4901 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
4902</div>
4903
4904<div class="doc_text">
4905
4906<h5>Syntax:</h5>
4907<pre>
4908 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
4909</pre>
4910
4911<h5>Overview:</h5>
4912
4913<p>
4914The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
4915target-specific value indicating the return address of the current function
4916or one of its callers.
4917</p>
4918
4919<h5>Arguments:</h5>
4920
4921<p>
4922The argument to this intrinsic indicates which function to return the address
4923for. Zero indicates the calling function, one indicates its caller, etc. The
4924argument is <b>required</b> to be a constant integer value.
4925</p>
4926
4927<h5>Semantics:</h5>
4928
4929<p>
4930The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
4931the return address of the specified call frame, or zero if it cannot be
4932identified. The value returned by this intrinsic is likely to be incorrect or 0
4933for arguments other than zero, so it should only be used for debugging purposes.
4934</p>
4935
4936<p>
4937Note that calling this intrinsic does not prevent function inlining or other
4938aggressive transformations, so the value returned may not be that of the obvious
4939source-language caller.
4940</p>
4941</div>
4942
4943
4944<!-- _______________________________________________________________________ -->
4945<div class="doc_subsubsection">
4946 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
4947</div>
4948
4949<div class="doc_text">
4950
4951<h5>Syntax:</h5>
4952<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4953 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4954</pre>
4955
4956<h5>Overview:</h5>
4957
4958<p>
4959The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
4960target-specific frame pointer value for the specified stack frame.
4961</p>
4962
4963<h5>Arguments:</h5>
4964
4965<p>
4966The argument to this intrinsic indicates which function to return the frame
4967pointer for. Zero indicates the calling function, one indicates its caller,
4968etc. The argument is <b>required</b> to be a constant integer value.
4969</p>
4970
4971<h5>Semantics:</h5>
4972
4973<p>
4974The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
4975the frame address of the specified call frame, or zero if it cannot be
4976identified. The value returned by this intrinsic is likely to be incorrect or 0
4977for arguments other than zero, so it should only be used for debugging purposes.
4978</p>
4979
4980<p>
4981Note that calling this intrinsic does not prevent function inlining or other
4982aggressive transformations, so the value returned may not be that of the obvious
4983source-language caller.
4984</p>
4985</div>
4986
4987<!-- _______________________________________________________________________ -->
4988<div class="doc_subsubsection">
4989 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
4990</div>
4991
4992<div class="doc_text">
4993
4994<h5>Syntax:</h5>
4995<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4996 declare i8 *@llvm.stacksave()
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4997</pre>
4998
4999<h5>Overview:</h5>
5000
5001<p>
5002The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
5003the function stack, for use with <a href="#int_stackrestore">
5004<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
5005features like scoped automatic variable sized arrays in C99.
5006</p>
5007
5008<h5>Semantics:</h5>
5009
5010<p>
5011This intrinsic returns a opaque pointer value that can be passed to <a
5012href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
5013<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
5014<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
5015state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
5016practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
5017that were allocated after the <tt>llvm.stacksave</tt> was executed.
5018</p>
5019
5020</div>
5021
5022<!-- _______________________________________________________________________ -->
5023<div class="doc_subsubsection">
5024 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
5025</div>
5026
5027<div class="doc_text">
5028
5029<h5>Syntax:</h5>
5030<pre>
5031 declare void @llvm.stackrestore(i8 * %ptr)
5032</pre>
5033
5034<h5>Overview:</h5>
5035
5036<p>
5037The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
5038the function stack to the state it was in when the corresponding <a
5039href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
5040useful for implementing language features like scoped automatic variable sized
5041arrays in C99.
5042</p>
5043
5044<h5>Semantics:</h5>
5045
5046<p>
5047See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
5048</p>
5049
5050</div>
5051
5052
5053<!-- _______________________________________________________________________ -->
5054<div class="doc_subsubsection">
5055 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
5056</div>
5057
5058<div class="doc_text">
5059
5060<h5>Syntax:</h5>
5061<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5062 declare void @llvm.prefetch(i8* &lt;address&gt;, i32 &lt;rw&gt;, i32 &lt;locality&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5063</pre>
5064
5065<h5>Overview:</h5>
5066
5067
5068<p>
5069The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
5070a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
5071no
5072effect on the behavior of the program but can change its performance
5073characteristics.
5074</p>
5075
5076<h5>Arguments:</h5>
5077
5078<p>
5079<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
5080determining if the fetch should be for a read (0) or write (1), and
5081<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
5082locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
5083<tt>locality</tt> arguments must be constant integers.
5084</p>
5085
5086<h5>Semantics:</h5>
5087
5088<p>
5089This intrinsic does not modify the behavior of the program. In particular,
5090prefetches cannot trap and do not produce a value. On targets that support this
5091intrinsic, the prefetch can provide hints to the processor cache for better
5092performance.
5093</p>
5094
5095</div>
5096
5097<!-- _______________________________________________________________________ -->
5098<div class="doc_subsubsection">
5099 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
5100</div>
5101
5102<div class="doc_text">
5103
5104<h5>Syntax:</h5>
5105<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]5106 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5107</pre>
5108
5109<h5>Overview:</h5>
5110
5111
5112<p>
5113The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]5114(PC) in a region of
5115code to simulators and other tools. The method is target specific, but it is
5116expected that the marker will use exported symbols to transmit the PC of the
5117marker.
5118The marker makes no guarantees that it will remain with any specific instruction
5119after optimizations. It is possible that the presence of a marker will inhibit
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5120optimizations. The intended use is to be inserted after optimizations to allow
5121correlations of simulation runs.
5122</p>
5123
5124<h5>Arguments:</h5>
5125
5126<p>
5127<tt>id</tt> is a numerical id identifying the marker.
5128</p>
5129
5130<h5>Semantics:</h5>
5131
5132<p>
5133This intrinsic does not modify the behavior of the program. Backends that do not
5134support this intrinisic may ignore it.
5135</p>
5136
5137</div>
5138
5139<!-- _______________________________________________________________________ -->
5140<div class="doc_subsubsection">
5141 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
5142</div>
5143
5144<div class="doc_text">
5145
5146<h5>Syntax:</h5>
5147<pre>
5148 declare i64 @llvm.readcyclecounter( )
5149</pre>
5150
5151<h5>Overview:</h5>
5152
5153
5154<p>
5155The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
5156counter register (or similar low latency, high accuracy clocks) on those targets
5157that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
5158As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
5159should only be used for small timings.
5160</p>
5161
5162<h5>Semantics:</h5>
5163
5164<p>
5165When directly supported, reading the cycle counter should not modify any memory.
5166Implementations are allowed to either return a application specific value or a
5167system wide value. On backends without support, this is lowered to a constant 0.
5168</p>
5169
5170</div>
5171
5172<!-- ======================================================================= -->
5173<div class="doc_subsection">
5174 <a name="int_libc">Standard C Library Intrinsics</a>
5175</div>
5176
5177<div class="doc_text">
5178<p>
5179LLVM provides intrinsics for a few important standard C library functions.
5180These intrinsics allow source-language front-ends to pass information about the
5181alignment of the pointer arguments to the code generator, providing opportunity
5182for more efficient code generation.
5183</p>
5184
5185</div>
5186
5187<!-- _______________________________________________________________________ -->
5188<div class="doc_subsubsection">
5189 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
5190</div>
5191
5192<div class="doc_text">
5193
5194<h5>Syntax:</h5>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5195<p>This is an overloaded intrinsic. You can use llvm.memcpy on any integer bit
5196width. Not all targets support all bit widths however.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5197<pre>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5198 declare void @llvm.memcpy.i8(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5199 i8 &lt;len&gt;, i32 &lt;align&gt;)
5200 declare void @llvm.memcpy.i16(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5201 i16 &lt;len&gt;, i32 &lt;align&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5202 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5203 i32 &lt;len&gt;, i32 &lt;align&gt;)
5204 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5205 i64 &lt;len&gt;, i32 &lt;align&gt;)
5206</pre>
5207
5208<h5>Overview:</h5>
5209
5210<p>
5211The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
5212location to the destination location.
5213</p>
5214
5215<p>
5216Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
5217intrinsics do not return a value, and takes an extra alignment argument.
5218</p>
5219
5220<h5>Arguments:</h5>
5221
5222<p>
5223The first argument is a pointer to the destination, the second is a pointer to
5224the source. The third argument is an integer argument
5225specifying the number of bytes to copy, and the fourth argument is the alignment
5226of the source and destination locations.
5227</p>
5228
5229<p>
5230If the call to this intrinisic has an alignment value that is not 0 or 1, then
5231the caller guarantees that both the source and destination pointers are aligned
5232to that boundary.
5233</p>
5234
5235<h5>Semantics:</h5>
5236
5237<p>
5238The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
5239location to the destination location, which are not allowed to overlap. It
5240copies "len" bytes of memory over. If the argument is known to be aligned to
5241some boundary, this can be specified as the fourth argument, otherwise it should
5242be set to 0 or 1.
5243</p>
5244</div>
5245
5246
5247<!-- _______________________________________________________________________ -->
5248<div class="doc_subsubsection">
5249 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
5250</div>
5251
5252<div class="doc_text">
5253
5254<h5>Syntax:</h5>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5255<p>This is an overloaded intrinsic. You can use llvm.memmove on any integer bit
5256width. Not all targets support all bit widths however.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5257<pre>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5258 declare void @llvm.memmove.i8(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5259 i8 &lt;len&gt;, i32 &lt;align&gt;)
5260 declare void @llvm.memmove.i16(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5261 i16 &lt;len&gt;, i32 &lt;align&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5262 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5263 i32 &lt;len&gt;, i32 &lt;align&gt;)
5264 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5265 i64 &lt;len&gt;, i32 &lt;align&gt;)
5266</pre>
5267
5268<h5>Overview:</h5>
5269
5270<p>
5271The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
5272location to the destination location. It is similar to the
Chris Lattnerdba16ea2008-01-06 19:51:52 +0000[diff] [blame]5273'<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to overlap.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5274</p>
5275
5276<p>
5277Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
5278intrinsics do not return a value, and takes an extra alignment argument.
5279</p>
5280
5281<h5>Arguments:</h5>
5282
5283<p>
5284The first argument is a pointer to the destination, the second is a pointer to
5285the source. The third argument is an integer argument
5286specifying the number of bytes to copy, and the fourth argument is the alignment
5287of the source and destination locations.
5288</p>
5289
5290<p>
5291If the call to this intrinisic has an alignment value that is not 0 or 1, then
5292the caller guarantees that the source and destination pointers are aligned to
5293that boundary.
5294</p>
5295
5296<h5>Semantics:</h5>
5297
5298<p>
5299The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
5300location to the destination location, which may overlap. It
5301copies "len" bytes of memory over. If the argument is known to be aligned to
5302some boundary, this can be specified as the fourth argument, otherwise it should
5303be set to 0 or 1.
5304</p>
5305</div>
5306
5307
5308<!-- _______________________________________________________________________ -->
5309<div class="doc_subsubsection">
5310 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
5311</div>
5312
5313<div class="doc_text">
5314
5315<h5>Syntax:</h5>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5316<p>This is an overloaded intrinsic. You can use llvm.memset on any integer bit
5317width. Not all targets support all bit widths however.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5318<pre>
Chris Lattner82c2e432008-11-21 16:42:48 +0000[diff] [blame]5319 declare void @llvm.memset.i8(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5320 i8 &lt;len&gt;, i32 &lt;align&gt;)
5321 declare void @llvm.memset.i16(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5322 i16 &lt;len&gt;, i32 &lt;align&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5323 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5324 i32 &lt;len&gt;, i32 &lt;align&gt;)
5325 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5326 i64 &lt;len&gt;, i32 &lt;align&gt;)
5327</pre>
5328
5329<h5>Overview:</h5>
5330
5331<p>
5332The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
5333byte value.
5334</p>
5335
5336<p>
5337Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
5338does not return a value, and takes an extra alignment argument.
5339</p>
5340
5341<h5>Arguments:</h5>
5342
5343<p>
5344The first argument is a pointer to the destination to fill, the second is the
5345byte value to fill it with, the third argument is an integer
5346argument specifying the number of bytes to fill, and the fourth argument is the
5347known alignment of destination location.
5348</p>
5349
5350<p>
5351If the call to this intrinisic has an alignment value that is not 0 or 1, then
5352the caller guarantees that the destination pointer is aligned to that boundary.
5353</p>
5354
5355<h5>Semantics:</h5>
5356
5357<p>
5358The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
5359the
5360destination location. If the argument is known to be aligned to some boundary,
5361this can be specified as the fourth argument, otherwise it should be set to 0 or
53621.
5363</p>
5364</div>
5365
5366
5367<!-- _______________________________________________________________________ -->
5368<div class="doc_subsubsection">
5369 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
5370</div>
5371
5372<div class="doc_text">
5373
5374<h5>Syntax:</h5>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5375<p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5376floating point or vector of floating point type. Not all targets support all
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5377types however.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5378<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5379 declare float @llvm.sqrt.f32(float %Val)
5380 declare double @llvm.sqrt.f64(double %Val)
5381 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
5382 declare fp128 @llvm.sqrt.f128(fp128 %Val)
5383 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5384</pre>
5385
5386<h5>Overview:</h5>
5387
5388<p>
5389The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5390returning the same value as the libm '<tt>sqrt</tt>' functions would. Unlike
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5391<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
Chris Lattnerf914a002008-01-29 07:00:44 +0000[diff] [blame]5392negative numbers other than -0.0 (which allows for better optimization, because
5393there is no need to worry about errno being set). <tt>llvm.sqrt(-0.0)</tt> is
5394defined to return -0.0 like IEEE sqrt.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5395</p>
5396
5397<h5>Arguments:</h5>
5398
5399<p>
5400The argument and return value are floating point numbers of the same type.
5401</p>
5402
5403<h5>Semantics:</h5>
5404
5405<p>
5406This function returns the sqrt of the specified operand if it is a nonnegative
5407floating point number.
5408</p>
5409</div>
5410
5411<!-- _______________________________________________________________________ -->
5412<div class="doc_subsubsection">
5413 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
5414</div>
5415
5416<div class="doc_text">
5417
5418<h5>Syntax:</h5>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5419<p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5420floating point or vector of floating point type. Not all targets support all
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5421types however.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5422<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5423 declare float @llvm.powi.f32(float %Val, i32 %power)
5424 declare double @llvm.powi.f64(double %Val, i32 %power)
5425 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
5426 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
5427 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5428</pre>
5429
5430<h5>Overview:</h5>
5431
5432<p>
5433The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
5434specified (positive or negative) power. The order of evaluation of
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5435multiplications is not defined. When a vector of floating point type is
5436used, the second argument remains a scalar integer value.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5437</p>
5438
5439<h5>Arguments:</h5>
5440
5441<p>
5442The second argument is an integer power, and the first is a value to raise to
5443that power.
5444</p>
5445
5446<h5>Semantics:</h5>
5447
5448<p>
5449This function returns the first value raised to the second power with an
5450unspecified sequence of rounding operations.</p>
5451</div>
5452
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5453<!-- _______________________________________________________________________ -->
5454<div class="doc_subsubsection">
5455 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
5456</div>
5457
5458<div class="doc_text">
5459
5460<h5>Syntax:</h5>
5461<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
5462floating point or vector of floating point type. Not all targets support all
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5463types however.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5464<pre>
5465 declare float @llvm.sin.f32(float %Val)
5466 declare double @llvm.sin.f64(double %Val)
5467 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
5468 declare fp128 @llvm.sin.f128(fp128 %Val)
5469 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
5470</pre>
5471
5472<h5>Overview:</h5>
5473
5474<p>
5475The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.
5476</p>
5477
5478<h5>Arguments:</h5>
5479
5480<p>
5481The argument and return value are floating point numbers of the same type.
5482</p>
5483
5484<h5>Semantics:</h5>
5485
5486<p>
5487This function returns the sine of the specified operand, returning the
5488same values as the libm <tt>sin</tt> functions would, and handles error
Dan Gohmaneaba92e2007-10-17 18:05:13 +0000[diff] [blame]5489conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5490</div>
5491
5492<!-- _______________________________________________________________________ -->
5493<div class="doc_subsubsection">
5494 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
5495</div>
5496
5497<div class="doc_text">
5498
5499<h5>Syntax:</h5>
5500<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
5501floating point or vector of floating point type. Not all targets support all
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5502types however.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5503<pre>
5504 declare float @llvm.cos.f32(float %Val)
5505 declare double @llvm.cos.f64(double %Val)
5506 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
5507 declare fp128 @llvm.cos.f128(fp128 %Val)
5508 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
5509</pre>
5510
5511<h5>Overview:</h5>
5512
5513<p>
5514The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.
5515</p>
5516
5517<h5>Arguments:</h5>
5518
5519<p>
5520The argument and return value are floating point numbers of the same type.
5521</p>
5522
5523<h5>Semantics:</h5>
5524
5525<p>
5526This function returns the cosine of the specified operand, returning the
5527same values as the libm <tt>cos</tt> functions would, and handles error
Dan Gohmaneaba92e2007-10-17 18:05:13 +0000[diff] [blame]5528conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5529</div>
5530
5531<!-- _______________________________________________________________________ -->
5532<div class="doc_subsubsection">
5533 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
5534</div>
5535
5536<div class="doc_text">
5537
5538<h5>Syntax:</h5>
5539<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
5540floating point or vector of floating point type. Not all targets support all
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5541types however.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5542<pre>
5543 declare float @llvm.pow.f32(float %Val, float %Power)
5544 declare double @llvm.pow.f64(double %Val, double %Power)
5545 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
5546 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
5547 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
5548</pre>
5549
5550<h5>Overview:</h5>
5551
5552<p>
5553The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
5554specified (positive or negative) power.
5555</p>
5556
5557<h5>Arguments:</h5>
5558
5559<p>
5560The second argument is a floating point power, and the first is a value to
5561raise to that power.
5562</p>
5563
5564<h5>Semantics:</h5>
5565
5566<p>
5567This function returns the first value raised to the second power,
5568returning the
5569same values as the libm <tt>pow</tt> functions would, and handles error
Dan Gohmaneaba92e2007-10-17 18:05:13 +0000[diff] [blame]5570conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5571</div>
5572
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5573
5574<!-- ======================================================================= -->
5575<div class="doc_subsection">
5576 <a name="int_manip">Bit Manipulation Intrinsics</a>
5577</div>
5578
5579<div class="doc_text">
5580<p>
5581LLVM provides intrinsics for a few important bit manipulation operations.
5582These allow efficient code generation for some algorithms.
5583</p>
5584
5585</div>
5586
5587<!-- _______________________________________________________________________ -->
5588<div class="doc_subsubsection">
5589 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
5590</div>
5591
5592<div class="doc_text">
5593
5594<h5>Syntax:</h5>
5595<p>This is an overloaded intrinsic function. You can use bswap on any integer
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5596type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5597<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5598 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
5599 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
5600 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5601</pre>
5602
5603<h5>Overview:</h5>
5604
5605<p>
5606The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
5607values with an even number of bytes (positive multiple of 16 bits). These are
5608useful for performing operations on data that is not in the target's native
5609byte order.
5610</p>
5611
5612<h5>Semantics:</h5>
5613
5614<p>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5615The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5616and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
5617intrinsic returns an i32 value that has the four bytes of the input i32
5618swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5619i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48</tt>,
5620<tt>llvm.bswap.i64</tt> and other intrinsics extend this concept to
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5621additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
5622</p>
5623
5624</div>
5625
5626<!-- _______________________________________________________________________ -->
5627<div class="doc_subsubsection">
5628 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
5629</div>
5630
5631<div class="doc_text">
5632
5633<h5>Syntax:</h5>
5634<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5635width. Not all targets support all bit widths however.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5636<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5637 declare i8 @llvm.ctpop.i8 (i8 &lt;src&gt;)
5638 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5639 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5640 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
5641 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5642</pre>
5643
5644<h5>Overview:</h5>
5645
5646<p>
5647The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
5648value.
5649</p>
5650
5651<h5>Arguments:</h5>
5652
5653<p>
5654The only argument is the value to be counted. The argument may be of any
5655integer type. The return type must match the argument type.
5656</p>
5657
5658<h5>Semantics:</h5>
5659
5660<p>
5661The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
5662</p>
5663</div>
5664
5665<!-- _______________________________________________________________________ -->
5666<div class="doc_subsubsection">
5667 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
5668</div>
5669
5670<div class="doc_text">
5671
5672<h5>Syntax:</h5>
5673<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5674integer bit width. Not all targets support all bit widths however.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5675<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5676 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
5677 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5678 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5679 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
5680 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5681</pre>
5682
5683<h5>Overview:</h5>
5684
5685<p>
5686The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
5687leading zeros in a variable.
5688</p>
5689
5690<h5>Arguments:</h5>
5691
5692<p>
5693The only argument is the value to be counted. The argument may be of any
5694integer type. The return type must match the argument type.
5695</p>
5696
5697<h5>Semantics:</h5>
5698
5699<p>
5700The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
5701in a variable. If the src == 0 then the result is the size in bits of the type
5702of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
5703</p>
5704</div>
5705
5706
5707
5708<!-- _______________________________________________________________________ -->
5709<div class="doc_subsubsection">
5710 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
5711</div>
5712
5713<div class="doc_text">
5714
5715<h5>Syntax:</h5>
5716<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5717integer bit width. Not all targets support all bit widths however.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5718<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5719 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
5720 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5721 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5722 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
5723 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5724</pre>
5725
5726<h5>Overview:</h5>
5727
5728<p>
5729The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
5730trailing zeros.
5731</p>
5732
5733<h5>Arguments:</h5>
5734
5735<p>
5736The only argument is the value to be counted. The argument may be of any
5737integer type. The return type must match the argument type.
5738</p>
5739
5740<h5>Semantics:</h5>
5741
5742<p>
5743The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
5744in a variable. If the src == 0 then the result is the size in bits of the type
5745of src. For example, <tt>llvm.cttz(2) = 1</tt>.
5746</p>
5747</div>
5748
5749<!-- _______________________________________________________________________ -->
5750<div class="doc_subsubsection">
5751 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
5752</div>
5753
5754<div class="doc_text">
5755
5756<h5>Syntax:</h5>
5757<p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5758on any integer bit width.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5759<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5760 declare i17 @llvm.part.select.i17 (i17 %val, i32 %loBit, i32 %hiBit)
5761 declare i29 @llvm.part.select.i29 (i29 %val, i32 %loBit, i32 %hiBit)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5762</pre>
5763
5764<h5>Overview:</h5>
5765<p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
5766range of bits from an integer value and returns them in the same bit width as
5767the original value.</p>
5768
5769<h5>Arguments:</h5>
5770<p>The first argument, <tt>%val</tt> and the result may be integer types of
5771any bit width but they must have the same bit width. The second and third
5772arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
5773
5774<h5>Semantics:</h5>
5775<p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
5776of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
5777<tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
5778operates in forward mode.</p>
5779<p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
5780right by <tt>%loBit</tt> bits and then ANDing it with a mask with
5781only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
5782<ol>
5783 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
5784 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
5785 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
5786 to determine the number of bits to retain.</li>
5787 <li>A mask of the retained bits is created by shifting a -1 value.</li>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5788 <li>The mask is ANDed with <tt>%val</tt> to produce the result.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5789</ol>
5790<p>In reverse mode, a similar computation is made except that the bits are
5791returned in the reverse order. So, for example, if <tt>X</tt> has the value
5792<tt>i16 0x0ACF (101011001111)</tt> and we apply
5793<tt>part.select(i16 X, 8, 3)</tt> to it, we get back the value
5794<tt>i16 0x0026 (000000100110)</tt>.</p>
5795</div>
5796
5797<div class="doc_subsubsection">
5798 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
5799</div>
5800
5801<div class="doc_text">
5802
5803<h5>Syntax:</h5>
5804<p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5805on any integer bit width.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5806<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5807 declare i17 @llvm.part.set.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
5808 declare i29 @llvm.part.set.i29.i9 (i29 %val, i9 %repl, i32 %lo, i32 %hi)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5809</pre>
5810
5811<h5>Overview:</h5>
5812<p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
5813of bits in an integer value with another integer value. It returns the integer
5814with the replaced bits.</p>
5815
5816<h5>Arguments:</h5>
5817<p>The first argument, <tt>%val</tt> and the result may be integer types of
5818any bit width but they must have the same bit width. <tt>%val</tt> is the value
5819whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
5820integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
5821type since they specify only a bit index.</p>
5822
5823<h5>Semantics:</h5>
5824<p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
5825of operation: forwards and reverse. If <tt>%lo</tt> is greater than
5826<tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
5827operates in forward mode.</p>
5828<p>For both modes, the <tt>%repl</tt> value is prepared for use by either
5829truncating it down to the size of the replacement area or zero extending it
5830up to that size.</p>
5831<p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
5832are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
5833in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5834to the <tt>%hi</tt>th bit.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5835<p>In reverse mode, a similar computation is made except that the bits are
5836reversed. That is, the <tt>0</tt>th bit in <tt>%repl</tt> replaces the
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]5837<tt>%hi</tt> bit in <tt>%val</tt> and etc. down to the <tt>%lo</tt>th bit.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5838<h5>Examples:</h5>
5839<pre>
5840 llvm.part.set(0xFFFF, 0, 4, 7) -&gt; 0xFF0F
5841 llvm.part.set(0xFFFF, 0, 7, 4) -&gt; 0xFF0F
5842 llvm.part.set(0xFFFF, 1, 7, 4) -&gt; 0xFF8F
5843 llvm.part.set(0xFFFF, F, 8, 3) -&gt; 0xFFE7
5844 llvm.part.set(0xFFFF, 0, 3, 8) -&gt; 0xFE07
5845</pre>
5846</div>
5847
5848<!-- ======================================================================= -->
5849<div class="doc_subsection">
5850 <a name="int_debugger">Debugger Intrinsics</a>
5851</div>
5852
5853<div class="doc_text">
5854<p>
5855The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
5856are described in the <a
5857href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
5858Debugging</a> document.
5859</p>
5860</div>
5861
5862
5863<!-- ======================================================================= -->
5864<div class="doc_subsection">
5865 <a name="int_eh">Exception Handling Intrinsics</a>
5866</div>
5867
5868<div class="doc_text">
5869<p> The LLVM exception handling intrinsics (which all start with
5870<tt>llvm.eh.</tt> prefix), are described in the <a
5871href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
5872Handling</a> document. </p>
5873</div>
5874
5875<!-- ======================================================================= -->
5876<div class="doc_subsection">
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5877 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5878</div>
5879
5880<div class="doc_text">
5881<p>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5882 This intrinsic makes it possible to excise one parameter, marked with
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5883 the <tt>nest</tt> attribute, from a function. The result is a callable
5884 function pointer lacking the nest parameter - the caller does not need
5885 to provide a value for it. Instead, the value to use is stored in
5886 advance in a "trampoline", a block of memory usually allocated
5887 on the stack, which also contains code to splice the nest value into the
5888 argument list. This is used to implement the GCC nested function address
5889 extension.
5890</p>
5891<p>
5892 For example, if the function is
5893 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
Bill Wendlinge262b4c2007-09-22 09:23:55 +0000[diff] [blame]5894 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as follows:</p>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5895<pre>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5896 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
5897 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
5898 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
5899 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5900</pre>
Bill Wendlinge262b4c2007-09-22 09:23:55 +0000[diff] [blame]5901 <p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
5902 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5903</div>
5904
5905<!-- _______________________________________________________________________ -->
5906<div class="doc_subsubsection">
5907 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
5908</div>
5909<div class="doc_text">
5910<h5>Syntax:</h5>
5911<pre>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5912declare i8* @llvm.init.trampoline(i8* &lt;tramp&gt;, i8* &lt;func&gt;, i8* &lt;nval&gt;)
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5913</pre>
5914<h5>Overview:</h5>
5915<p>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5916 This fills the memory pointed to by <tt>tramp</tt> with code
5917 and returns a function pointer suitable for executing it.
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5918</p>
5919<h5>Arguments:</h5>
5920<p>
5921 The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
5922 pointers. The <tt>tramp</tt> argument must point to a sufficiently large
5923 and sufficiently aligned block of memory; this memory is written to by the
Duncan Sands35012212007-08-22 23:39:54 +0000[diff] [blame]5924 intrinsic. Note that the size and the alignment are target-specific - LLVM
5925 currently provides no portable way of determining them, so a front-end that
5926 generates this intrinsic needs to have some target-specific knowledge.
5927 The <tt>func</tt> argument must hold a function bitcast to an <tt>i8*</tt>.
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5928</p>
5929<h5>Semantics:</h5>
5930<p>
5931 The block of memory pointed to by <tt>tramp</tt> is filled with target
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5932 dependent code, turning it into a function. A pointer to this function is
5933 returned, but needs to be bitcast to an
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5934 <a href="#int_trampoline">appropriate function pointer type</a>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5935 before being called. The new function's signature is the same as that of
5936 <tt>func</tt> with any arguments marked with the <tt>nest</tt> attribute
5937 removed. At most one such <tt>nest</tt> argument is allowed, and it must be
5938 of pointer type. Calling the new function is equivalent to calling
5939 <tt>func</tt> with the same argument list, but with <tt>nval</tt> used for the
5940 missing <tt>nest</tt> argument. If, after calling
5941 <tt>llvm.init.trampoline</tt>, the memory pointed to by <tt>tramp</tt> is
5942 modified, then the effect of any later call to the returned function pointer is
5943 undefined.
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5944</p>
5945</div>
5946
5947<!-- ======================================================================= -->
5948<div class="doc_subsection">
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]5949 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
5950</div>
5951
5952<div class="doc_text">
5953<p>
5954 These intrinsic functions expand the "universal IR" of LLVM to represent
5955 hardware constructs for atomic operations and memory synchronization. This
5956 provides an interface to the hardware, not an interface to the programmer. It
Chris Lattner96451482008-08-05 18:29:16 +0000[diff] [blame]5957 is aimed at a low enough level to allow any programming models or APIs
5958 (Application Programming Interfaces) which
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]5959 need atomic behaviors to map cleanly onto it. It is also modeled primarily on
5960 hardware behavior. Just as hardware provides a "universal IR" for source
5961 languages, it also provides a starting point for developing a "universal"
5962 atomic operation and synchronization IR.
5963</p>
5964<p>
5965 These do <em>not</em> form an API such as high-level threading libraries,
5966 software transaction memory systems, atomic primitives, and intrinsic
5967 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
5968 application libraries. The hardware interface provided by LLVM should allow
5969 a clean implementation of all of these APIs and parallel programming models.
5970 No one model or paradigm should be selected above others unless the hardware
5971 itself ubiquitously does so.
5972
5973</p>
5974</div>
5975
5976<!-- _______________________________________________________________________ -->
5977<div class="doc_subsubsection">
5978 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
5979</div>
5980<div class="doc_text">
5981<h5>Syntax:</h5>
5982<pre>
5983declare void @llvm.memory.barrier( i1 &lt;ll&gt;, i1 &lt;ls&gt;, i1 &lt;sl&gt;, i1 &lt;ss&gt;,
5984i1 &lt;device&gt; )
5985
5986</pre>
5987<h5>Overview:</h5>
5988<p>
5989 The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
5990 specific pairs of memory access types.
5991</p>
5992<h5>Arguments:</h5>
5993<p>
5994 The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
5995 The first four arguments enables a specific barrier as listed below. The fith
5996 argument specifies that the barrier applies to io or device or uncached memory.
5997
5998</p>
5999 <ul>
6000 <li><tt>ll</tt>: load-load barrier</li>
6001 <li><tt>ls</tt>: load-store barrier</li>
6002 <li><tt>sl</tt>: store-load barrier</li>
6003 <li><tt>ss</tt>: store-store barrier</li>
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]6004 <li><tt>device</tt>: barrier applies to device and uncached memory also.</li>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6005 </ul>
6006<h5>Semantics:</h5>
6007<p>
6008 This intrinsic causes the system to enforce some ordering constraints upon
6009 the loads and stores of the program. This barrier does not indicate
6010 <em>when</em> any events will occur, it only enforces an <em>order</em> in
6011 which they occur. For any of the specified pairs of load and store operations
6012 (f.ex. load-load, or store-load), all of the first operations preceding the
6013 barrier will complete before any of the second operations succeeding the
6014 barrier begin. Specifically the semantics for each pairing is as follows:
6015</p>
6016 <ul>
6017 <li><tt>ll</tt>: All loads before the barrier must complete before any load
6018 after the barrier begins.</li>
6019
6020 <li><tt>ls</tt>: All loads before the barrier must complete before any
6021 store after the barrier begins.</li>
6022 <li><tt>ss</tt>: All stores before the barrier must complete before any
6023 store after the barrier begins.</li>
6024 <li><tt>sl</tt>: All stores before the barrier must complete before any
6025 load after the barrier begins.</li>
6026 </ul>
6027<p>
6028 These semantics are applied with a logical "and" behavior when more than one
6029 is enabled in a single memory barrier intrinsic.
6030</p>
6031<p>
6032 Backends may implement stronger barriers than those requested when they do not
6033 support as fine grained a barrier as requested. Some architectures do not
6034 need all types of barriers and on such architectures, these become noops.
6035</p>
6036<h5>Example:</h5>
6037<pre>
6038%ptr = malloc i32
6039 store i32 4, %ptr
6040
6041%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
6042 call void @llvm.memory.barrier( i1 false, i1 true, i1 false, i1 false )
6043 <i>; guarantee the above finishes</i>
6044 store i32 8, %ptr <i>; before this begins</i>
6045</pre>
6046</div>
6047
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6048<!-- _______________________________________________________________________ -->
6049<div class="doc_subsubsection">
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6050 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6051</div>
6052<div class="doc_text">
6053<h5>Syntax:</h5>
6054<p>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6055 This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on
6056 any integer bit width and for different address spaces. Not all targets
6057 support all bit widths however.</p>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6058
6059<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6060declare i8 @llvm.atomic.cmp.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt; )
6061declare i16 @llvm.atomic.cmp.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt; )
6062declare i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt; )
6063declare i64 @llvm.atomic.cmp.swap.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;cmp&gt;, i64 &lt;val&gt; )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6064
6065</pre>
6066<h5>Overview:</h5>
6067<p>
6068 This loads a value in memory and compares it to a given value. If they are
6069 equal, it stores a new value into the memory.
6070</p>
6071<h5>Arguments:</h5>
6072<p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6073 The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result as
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6074 well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
6075 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
6076 this integer type. While any bit width integer may be used, targets may only
6077 lower representations they support in hardware.
6078
6079</p>
6080<h5>Semantics:</h5>
6081<p>
6082 This entire intrinsic must be executed atomically. It first loads the value
6083 in memory pointed to by <tt>ptr</tt> and compares it with the value
6084 <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the memory. The
6085 loaded value is yielded in all cases. This provides the equivalent of an
6086 atomic compare-and-swap operation within the SSA framework.
6087</p>
6088<h5>Examples:</h5>
6089
6090<pre>
6091%ptr = malloc i32
6092 store i32 4, %ptr
6093
6094%val1 = add i32 4, 4
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6095%result1 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 4, %val1 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6096 <i>; yields {i32}:result1 = 4</i>
6097%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6098%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6099
6100%val2 = add i32 1, 1
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6101%result2 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 5, %val2 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6102 <i>; yields {i32}:result2 = 8</i>
6103%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
6104
6105%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
6106</pre>
6107</div>
6108
6109<!-- _______________________________________________________________________ -->
6110<div class="doc_subsubsection">
6111 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
6112</div>
6113<div class="doc_text">
6114<h5>Syntax:</h5>
6115
6116<p>
6117 This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
6118 integer bit width. Not all targets support all bit widths however.</p>
6119<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6120declare i8 @llvm.atomic.swap.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;val&gt; )
6121declare i16 @llvm.atomic.swap.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;val&gt; )
6122declare i32 @llvm.atomic.swap.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;val&gt; )
6123declare i64 @llvm.atomic.swap.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;val&gt; )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6124
6125</pre>
6126<h5>Overview:</h5>
6127<p>
6128 This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
6129 the value from memory. It then stores the value in <tt>val</tt> in the memory
6130 at <tt>ptr</tt>.
6131</p>
6132<h5>Arguments:</h5>
6133
6134<p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6135 The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both the
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6136 <tt>val</tt> argument and the result must be integers of the same bit width.
6137 The first argument, <tt>ptr</tt>, must be a pointer to a value of this
6138 integer type. The targets may only lower integer representations they
6139 support.
6140</p>
6141<h5>Semantics:</h5>
6142<p>
6143 This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
6144 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
6145 equivalent of an atomic swap operation within the SSA framework.
6146
6147</p>
6148<h5>Examples:</h5>
6149<pre>
6150%ptr = malloc i32
6151 store i32 4, %ptr
6152
6153%val1 = add i32 4, 4
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6154%result1 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val1 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6155 <i>; yields {i32}:result1 = 4</i>
6156%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6157%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6158
6159%val2 = add i32 1, 1
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6160%result2 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val2 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6161 <i>; yields {i32}:result2 = 8</i>
6162
6163%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
6164%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
6165</pre>
6166</div>
6167
6168<!-- _______________________________________________________________________ -->
6169<div class="doc_subsubsection">
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6170 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6171
6172</div>
6173<div class="doc_text">
6174<h5>Syntax:</h5>
6175<p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6176 This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on any
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6177 integer bit width. Not all targets support all bit widths however.</p>
6178<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6179declare i8 @llvm.atomic.load.add.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6180declare i16 @llvm.atomic.load.add.i16..p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6181declare i32 @llvm.atomic.load.add.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6182declare i64 @llvm.atomic.load.add.i64..p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6183
6184</pre>
6185<h5>Overview:</h5>
6186<p>
6187 This intrinsic adds <tt>delta</tt> to the value stored in memory at
6188 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.
6189</p>
6190<h5>Arguments:</h5>
6191<p>
6192
6193 The intrinsic takes two arguments, the first a pointer to an integer value
6194 and the second an integer value. The result is also an integer value. These
6195 integer types can have any bit width, but they must all have the same bit
6196 width. The targets may only lower integer representations they support.
6197</p>
6198<h5>Semantics:</h5>
6199<p>
6200 This intrinsic does a series of operations atomically. It first loads the
6201 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
6202 to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.
6203</p>
6204
6205<h5>Examples:</h5>
6206<pre>
6207%ptr = malloc i32
6208 store i32 4, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6209%result1 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 4 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6210 <i>; yields {i32}:result1 = 4</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6211%result2 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 2 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6212 <i>; yields {i32}:result2 = 8</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6213%result3 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 5 )
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6214 <i>; yields {i32}:result3 = 10</i>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6215%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]6216</pre>
6217</div>
6218
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6219<!-- _______________________________________________________________________ -->
6220<div class="doc_subsubsection">
6221 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
6222
6223</div>
6224<div class="doc_text">
6225<h5>Syntax:</h5>
6226<p>
6227 This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6228 any integer bit width and for different address spaces. Not all targets
6229 support all bit widths however.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6230<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6231declare i8 @llvm.atomic.load.sub.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6232declare i16 @llvm.atomic.load.sub.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6233declare i32 @llvm.atomic.load.sub.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6234declare i64 @llvm.atomic.load.sub.i64.p0i32( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6235
6236</pre>
6237<h5>Overview:</h5>
6238<p>
6239 This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
6240 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.
6241</p>
6242<h5>Arguments:</h5>
6243<p>
6244
6245 The intrinsic takes two arguments, the first a pointer to an integer value
6246 and the second an integer value. The result is also an integer value. These
6247 integer types can have any bit width, but they must all have the same bit
6248 width. The targets may only lower integer representations they support.
6249</p>
6250<h5>Semantics:</h5>
6251<p>
6252 This intrinsic does a series of operations atomically. It first loads the
6253 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
6254 result to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.
6255</p>
6256
6257<h5>Examples:</h5>
6258<pre>
6259%ptr = malloc i32
6260 store i32 8, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6261%result1 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 4 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6262 <i>; yields {i32}:result1 = 8</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6263%result2 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 2 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6264 <i>; yields {i32}:result2 = 4</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6265%result3 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 5 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6266 <i>; yields {i32}:result3 = 2</i>
6267%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
6268</pre>
6269</div>
6270
6271<!-- _______________________________________________________________________ -->
6272<div class="doc_subsubsection">
6273 <a name="int_atomic_load_and">'<tt>llvm.atomic.load.and.*</tt>' Intrinsic</a><br>
6274 <a name="int_atomic_load_nand">'<tt>llvm.atomic.load.nand.*</tt>' Intrinsic</a><br>
6275 <a name="int_atomic_load_or">'<tt>llvm.atomic.load.or.*</tt>' Intrinsic</a><br>
6276 <a name="int_atomic_load_xor">'<tt>llvm.atomic.load.xor.*</tt>' Intrinsic</a><br>
6277
6278</div>
6279<div class="doc_text">
6280<h5>Syntax:</h5>
6281<p>
6282 These are overloaded intrinsics. You can use <tt>llvm.atomic.load_and</tt>,
6283 <tt>llvm.atomic.load_nand</tt>, <tt>llvm.atomic.load_or</tt>, and
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6284 <tt>llvm.atomic.load_xor</tt> on any integer bit width and for different
6285 address spaces. Not all targets support all bit widths however.</p>
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6286<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6287declare i8 @llvm.atomic.load.and.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6288declare i16 @llvm.atomic.load.and.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6289declare i32 @llvm.atomic.load.and.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6290declare i64 @llvm.atomic.load.and.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6291
6292</pre>
6293
6294<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6295declare i8 @llvm.atomic.load.or.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6296declare i16 @llvm.atomic.load.or.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6297declare i32 @llvm.atomic.load.or.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6298declare i64 @llvm.atomic.load.or.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6299
6300</pre>
6301
6302<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6303declare i8 @llvm.atomic.load.nand.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6304declare i16 @llvm.atomic.load.nand.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6305declare i32 @llvm.atomic.load.nand.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6306declare i64 @llvm.atomic.load.nand.i64.p0i32( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6307
6308</pre>
6309
6310<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6311declare i8 @llvm.atomic.load.xor.i8.p0i32( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6312declare i16 @llvm.atomic.load.xor.i16.p0i32( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6313declare i32 @llvm.atomic.load.xor.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6314declare i64 @llvm.atomic.load.xor.i64.p0i32( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6315
6316</pre>
6317<h5>Overview:</h5>
6318<p>
6319 These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
6320 the value stored in memory at <tt>ptr</tt>. It yields the original value
6321 at <tt>ptr</tt>.
6322</p>
6323<h5>Arguments:</h5>
6324<p>
6325
6326 These intrinsics take two arguments, the first a pointer to an integer value
6327 and the second an integer value. The result is also an integer value. These
6328 integer types can have any bit width, but they must all have the same bit
6329 width. The targets may only lower integer representations they support.
6330</p>
6331<h5>Semantics:</h5>
6332<p>
6333 These intrinsics does a series of operations atomically. They first load the
6334 value stored at <tt>ptr</tt>. They then do the bitwise operation
6335 <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the original
6336 value stored at <tt>ptr</tt>.
6337</p>
6338
6339<h5>Examples:</h5>
6340<pre>
6341%ptr = malloc i32
6342 store i32 0x0F0F, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6343%result0 = call i32 @llvm.atomic.load.nand.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6344 <i>; yields {i32}:result0 = 0x0F0F</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6345%result1 = call i32 @llvm.atomic.load.and.i32.p0i32( i32* %ptr, i32 0xFF )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6346 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6347%result2 = call i32 @llvm.atomic.load.or.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6348 <i>; yields {i32}:result2 = 0xF0</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6349%result3 = call i32 @llvm.atomic.load.xor.i32.p0i32( i32* %ptr, i32 0F )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6350 <i>; yields {i32}:result3 = FF</i>
6351%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
6352</pre>
6353</div>
6354
6355
6356<!-- _______________________________________________________________________ -->
6357<div class="doc_subsubsection">
6358 <a name="int_atomic_load_max">'<tt>llvm.atomic.load.max.*</tt>' Intrinsic</a><br>
6359 <a name="int_atomic_load_min">'<tt>llvm.atomic.load.min.*</tt>' Intrinsic</a><br>
6360 <a name="int_atomic_load_umax">'<tt>llvm.atomic.load.umax.*</tt>' Intrinsic</a><br>
6361 <a name="int_atomic_load_umin">'<tt>llvm.atomic.load.umin.*</tt>' Intrinsic</a><br>
6362
6363</div>
6364<div class="doc_text">
6365<h5>Syntax:</h5>
6366<p>
6367 These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
6368 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6369 <tt>llvm.atomic.load_umin</tt> on any integer bit width and for different
6370 address spaces. Not all targets
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6371 support all bit widths however.</p>
6372<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6373declare i8 @llvm.atomic.load.max.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6374declare i16 @llvm.atomic.load.max.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6375declare i32 @llvm.atomic.load.max.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6376declare i64 @llvm.atomic.load.max.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6377
6378</pre>
6379
6380<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6381declare i8 @llvm.atomic.load.min.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6382declare i16 @llvm.atomic.load.min.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6383declare i32 @llvm.atomic.load.min.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6384declare i64 @llvm.atomic.load.min.i64..p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6385
6386</pre>
6387
6388<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6389declare i8 @llvm.atomic.load.umax.i8.p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6390declare i16 @llvm.atomic.load.umax.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6391declare i32 @llvm.atomic.load.umax.i32.p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6392declare i64 @llvm.atomic.load.umax.i64.p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6393
6394</pre>
6395
6396<pre>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6397declare i8 @llvm.atomic.load.umin.i8..p0i8( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
6398declare i16 @llvm.atomic.load.umin.i16.p0i16( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
6399declare i32 @llvm.atomic.load.umin.i32..p0i32( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
6400declare i64 @llvm.atomic.load.umin.i64..p0i64( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6401
6402</pre>
6403<h5>Overview:</h5>
6404<p>
6405 These intrinsics takes the signed or unsigned minimum or maximum of
6406 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
6407 original value at <tt>ptr</tt>.
6408</p>
6409<h5>Arguments:</h5>
6410<p>
6411
6412 These intrinsics take two arguments, the first a pointer to an integer value
6413 and the second an integer value. The result is also an integer value. These
6414 integer types can have any bit width, but they must all have the same bit
6415 width. The targets may only lower integer representations they support.
6416</p>
6417<h5>Semantics:</h5>
6418<p>
6419 These intrinsics does a series of operations atomically. They first load the
6420 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or max
6421 <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They yield
6422 the original value stored at <tt>ptr</tt>.
6423</p>
6424
6425<h5>Examples:</h5>
6426<pre>
6427%ptr = malloc i32
6428 store i32 7, %ptr
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6429%result0 = call i32 @llvm.atomic.load.min.i32.p0i32( i32* %ptr, i32 -2 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6430 <i>; yields {i32}:result0 = 7</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6431%result1 = call i32 @llvm.atomic.load.max.i32.p0i32( i32* %ptr, i32 8 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6432 <i>; yields {i32}:result1 = -2</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6433%result2 = call i32 @llvm.atomic.load.umin.i32.p0i32( i32* %ptr, i32 10 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6434 <i>; yields {i32}:result2 = 8</i>
Mon P Wangce3ac892008-07-30 04:36:53 +0000[diff] [blame]6435%result3 = call i32 @llvm.atomic.load.umax.i32.p0i32( i32* %ptr, i32 30 )
Mon P Wang6bde9ec2008-06-25 08:15:39 +0000[diff] [blame]6436 <i>; yields {i32}:result3 = 8</i>
6437%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
6438</pre>
6439</div>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]6440
6441<!-- ======================================================================= -->
6442<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6443 <a name="int_general">General Intrinsics</a>
6444</div>
6445
6446<div class="doc_text">
6447<p> This class of intrinsics is designed to be generic and has
6448no specific purpose. </p>
6449</div>
6450
6451<!-- _______________________________________________________________________ -->
6452<div class="doc_subsubsection">
6453 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
6454</div>
6455
6456<div class="doc_text">
6457
6458<h5>Syntax:</h5>
6459<pre>
6460 declare void @llvm.var.annotation(i8* &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6461</pre>
6462
6463<h5>Overview:</h5>
6464
6465<p>
6466The '<tt>llvm.var.annotation</tt>' intrinsic
6467</p>
6468
6469<h5>Arguments:</h5>
6470
6471<p>
6472The first argument is a pointer to a value, the second is a pointer to a
6473global string, the third is a pointer to a global string which is the source
6474file name, and the last argument is the line number.
6475</p>
6476
6477<h5>Semantics:</h5>
6478
6479<p>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]6480This intrinsic allows annotation of local variables with arbitrary strings.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6481This can be useful for special purpose optimizations that want to look for these
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]6482annotations. These have no other defined use, they are ignored by code
6483generation and optimization.
6484</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6485</div>
6486
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6487<!-- _______________________________________________________________________ -->
6488<div class="doc_subsubsection">
Tanya Lattnerc9869b12007-09-21 23:57:59 +0000[diff] [blame]6489 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6490</div>
6491
6492<div class="doc_text">
6493
6494<h5>Syntax:</h5>
Tanya Lattnere545be72007-09-21 23:56:27 +0000[diff] [blame]6495<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
6496any integer bit width.
6497</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6498<pre>
Tanya Lattner09161fe2007-09-22 00:03:01 +0000[diff] [blame]6499 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6500 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6501 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6502 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6503 declare i256 @llvm.annotation.i256(i256 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6504</pre>
6505
6506<h5>Overview:</h5>
Tanya Lattnere545be72007-09-21 23:56:27 +0000[diff] [blame]6507
6508<p>
6509The '<tt>llvm.annotation</tt>' intrinsic.
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6510</p>
6511
6512<h5>Arguments:</h5>
6513
6514<p>
6515The first argument is an integer value (result of some expression),
6516the second is a pointer to a global string, the third is a pointer to a global
6517string which is the source file name, and the last argument is the line number.
Tanya Lattnere545be72007-09-21 23:56:27 +0000[diff] [blame]6518It returns the value of the first argument.
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6519</p>
6520
6521<h5>Semantics:</h5>
6522
6523<p>
6524This intrinsic allows annotations to be put on arbitrary expressions
6525with arbitrary strings. This can be useful for special purpose optimizations
6526that want to look for these annotations. These have no other defined use, they
6527are ignored by code generation and optimization.
Dan Gohman2672f3e2008-10-14 16:51:45 +0000[diff] [blame]6528</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6529</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6530
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]6531<!-- _______________________________________________________________________ -->
6532<div class="doc_subsubsection">
6533 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
6534</div>
6535
6536<div class="doc_text">
6537
6538<h5>Syntax:</h5>
6539<pre>
6540 declare void @llvm.trap()
6541</pre>
6542
6543<h5>Overview:</h5>
6544
6545<p>
6546The '<tt>llvm.trap</tt>' intrinsic
6547</p>
6548
6549<h5>Arguments:</h5>
6550
6551<p>
6552None
6553</p>
6554
6555<h5>Semantics:</h5>
6556
6557<p>
6558This intrinsics is lowered to the target dependent trap instruction. If the
6559target does not have a trap instruction, this intrinsic will be lowered to the
6560call of the abort() function.
6561</p>
6562</div>
6563
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6564<!-- _______________________________________________________________________ -->
6565<div class="doc_subsubsection">
Misha Brukman5dd7f4d2008-11-22 23:55:29 +0000[diff] [blame]6566 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
Bill Wendlinge4164592008-11-19 05:56:17 +0000[diff] [blame]6567</div>
6568<div class="doc_text">
6569<h5>Syntax:</h5>
6570<pre>
6571declare void @llvm.stackprotector( i8* &lt;guard&gt;, i8** &lt;slot&gt; )
6572
6573</pre>
6574<h5>Overview:</h5>
6575<p>
6576 The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and stores
6577 it onto the stack at <tt>slot</tt>. The stack slot is adjusted to ensure that
6578 it is placed on the stack before local variables.
6579</p>
6580<h5>Arguments:</h5>
6581<p>
6582 The <tt>llvm.stackprotector</tt> intrinsic requires two pointer arguments. The
6583 first argument is the value loaded from the stack guard
6584 <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt> that
6585 has enough space to hold the value of the guard.
6586</p>
6587<h5>Semantics:</h5>
6588<p>
6589 This intrinsic causes the prologue/epilogue inserter to force the position of
6590 the <tt>AllocaInst</tt> stack slot to be before local variables on the
6591 stack. This is to ensure that if a local variable on the stack is overwritten,
6592 it will destroy the value of the guard. When the function exits, the guard on
6593 the stack is checked against the original guard. If they're different, then
6594 the program aborts by calling the <tt>__stack_chk_fail()</tt> function.
6595</p>
6596</div>
6597
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6598<!-- *********************************************************************** -->
6599<hr>
6600<address>
6601 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
Misha Brukman947321d2008-12-11 17:34:48 +0000[diff] [blame]6602 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6603 <a href="http://validator.w3.org/check/referer"><img
Misha Brukman947321d2008-12-11 17:34:48 +0000[diff] [blame]6604 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6605
6606 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
6607 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
6608 Last modified: $Date$
6609</address>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]6610
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6611</body>
6612</html>