Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 96e7e09f2642faf8f751be61c69d44687f217973 / . / docs / LangRef.html
blob: 25b9e026a26df68643b29309a8f8c25681b87020 [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>
25 <li><a href="#globalvars">Global Variables</a></li>
26 <li><a href="#functionstructure">Functions</a></li>
27 <li><a href="#aliasstructure">Aliases</a>
28 <li><a href="#paramattrs">Parameter Attributes</a></li>
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]29 <li><a href="#gc">Garbage Collector Names</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]30 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
31 <li><a href="#datalayout">Data Layout</a></li>
32 </ol>
33 </li>
34 <li><a href="#typesystem">Type System</a>
35 <ol>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]36 <li><a href="#t_classifications">Type Classifications</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]37 <li><a href="#t_primitive">Primitive Types</a>
38 <ol>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]39 <li><a href="#t_floating">Floating Point Types</a></li>
40 <li><a href="#t_void">Void Type</a></li>
41 <li><a href="#t_label">Label Type</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]42 </ol>
43 </li>
44 <li><a href="#t_derived">Derived Types</a>
45 <ol>
Chris Lattner251ab812007-12-18 06:18:21 +0000[diff] [blame]46 <li><a href="#t_integer">Integer Type</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]47 <li><a href="#t_array">Array Type</a></li>
48 <li><a href="#t_function">Function Type</a></li>
49 <li><a href="#t_pointer">Pointer Type</a></li>
50 <li><a href="#t_struct">Structure Type</a></li>
51 <li><a href="#t_pstruct">Packed Structure Type</a></li>
52 <li><a href="#t_vector">Vector Type</a></li>
53 <li><a href="#t_opaque">Opaque Type</a></li>
54 </ol>
55 </li>
56 </ol>
57 </li>
58 <li><a href="#constants">Constants</a>
59 <ol>
60 <li><a href="#simpleconstants">Simple Constants</a>
61 <li><a href="#aggregateconstants">Aggregate Constants</a>
62 <li><a href="#globalconstants">Global Variable and Function Addresses</a>
63 <li><a href="#undefvalues">Undefined Values</a>
64 <li><a href="#constantexprs">Constant Expressions</a>
65 </ol>
66 </li>
67 <li><a href="#othervalues">Other Values</a>
68 <ol>
69 <li><a href="#inlineasm">Inline Assembler Expressions</a>
70 </ol>
71 </li>
72 <li><a href="#instref">Instruction Reference</a>
73 <ol>
74 <li><a href="#terminators">Terminator Instructions</a>
75 <ol>
76 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
77 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
78 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
79 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
80 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
81 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
82 </ol>
83 </li>
84 <li><a href="#binaryops">Binary Operations</a>
85 <ol>
86 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
87 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
88 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
89 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
90 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
91 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
92 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
93 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
94 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
95 </ol>
96 </li>
97 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
98 <ol>
99 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
100 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
101 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
102 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
103 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
104 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
105 </ol>
106 </li>
107 <li><a href="#vectorops">Vector Operations</a>
108 <ol>
109 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
110 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
111 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
112 </ol>
113 </li>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]114 <li><a href="#aggregateops">Aggregate Operations</a>
115 <ol>
116 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
117 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
118 </ol>
119 </li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]120 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
121 <ol>
122 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
123 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
124 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
125 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
126 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
127 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
128 </ol>
129 </li>
130 <li><a href="#convertops">Conversion Operations</a>
131 <ol>
132 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
133 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
134 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
135 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
136 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
137 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
138 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
139 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
140 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
141 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
142 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
143 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
144 </ol>
145 <li><a href="#otherops">Other Operations</a>
146 <ol>
147 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
148 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]149 <li><a href="#i_vicmp">'<tt>vicmp</tt>' Instruction</a></li>
150 <li><a href="#i_vfcmp">'<tt>vfcmp</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]151 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
152 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
153 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
154 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]155 <li><a href="#i_getresult">'<tt>getresult</tt>' Instruction</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]156 </ol>
157 </li>
158 </ol>
159 </li>
160 <li><a href="#intrinsics">Intrinsic Functions</a>
161 <ol>
162 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
163 <ol>
164 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
165 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
166 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
167 </ol>
168 </li>
169 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
170 <ol>
171 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
172 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
173 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
174 </ol>
175 </li>
176 <li><a href="#int_codegen">Code Generator Intrinsics</a>
177 <ol>
178 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
179 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
180 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
181 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
182 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
183 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
184 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
185 </ol>
186 </li>
187 <li><a href="#int_libc">Standard C Library Intrinsics</a>
188 <ol>
189 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
190 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
191 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
192 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
193 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]194 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
195 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
196 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]197 </ol>
198 </li>
199 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
200 <ol>
201 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
202 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
203 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
204 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
205 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
206 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
207 </ol>
208 </li>
209 <li><a href="#int_debugger">Debugger intrinsics</a></li>
210 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]211 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]212 <ol>
213 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]214 </ol>
215 </li>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]216 <li><a href="#int_atomics">Atomic intrinsics</a>
217 <ol>
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]218 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
219 <li><a href="#int_atomic_lcs"><tt>llvm.atomic.lcs</tt></a></li>
220 <li><a href="#int_atomic_las"><tt>llvm.atomic.las</tt></a></li>
221 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]222 </ol>
223 </li>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]224 <li><a href="#int_general">General intrinsics</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]225 <ol>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]226 <li><a href="#int_var_annotation">
Tanya Lattner51369f32007-09-22 00:01:26 +0000[diff] [blame]227 <tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]228 <li><a href="#int_annotation">
Tanya Lattner51369f32007-09-22 00:01:26 +0000[diff] [blame]229 <tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]230 <li><a href="#int_trap">
231 <tt>llvm.trap</tt>' Intrinsic</a></li>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]232 </ol>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]233 </li>
234 </ol>
235 </li>
236</ol>
237
238<div class="doc_author">
239 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
240 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
241</div>
242
243<!-- *********************************************************************** -->
244<div class="doc_section"> <a name="abstract">Abstract </a></div>
245<!-- *********************************************************************** -->
246
247<div class="doc_text">
248<p>This document is a reference manual for the LLVM assembly language.
249LLVM is an SSA based representation that provides type safety,
250low-level operations, flexibility, and the capability of representing
251'all' high-level languages cleanly. It is the common code
252representation used throughout all phases of the LLVM compilation
253strategy.</p>
254</div>
255
256<!-- *********************************************************************** -->
257<div class="doc_section"> <a name="introduction">Introduction</a> </div>
258<!-- *********************************************************************** -->
259
260<div class="doc_text">
261
262<p>The LLVM code representation is designed to be used in three
263different forms: as an in-memory compiler IR, as an on-disk bitcode
264representation (suitable for fast loading by a Just-In-Time compiler),
265and as a human readable assembly language representation. This allows
266LLVM to provide a powerful intermediate representation for efficient
267compiler transformations and analysis, while providing a natural means
268to debug and visualize the transformations. The three different forms
269of LLVM are all equivalent. This document describes the human readable
270representation and notation.</p>
271
272<p>The LLVM representation aims to be light-weight and low-level
273while being expressive, typed, and extensible at the same time. It
274aims to be a "universal IR" of sorts, by being at a low enough level
275that high-level ideas may be cleanly mapped to it (similar to how
276microprocessors are "universal IR's", allowing many source languages to
277be mapped to them). By providing type information, LLVM can be used as
278the target of optimizations: for example, through pointer analysis, it
279can be proven that a C automatic variable is never accessed outside of
280the current function... allowing it to be promoted to a simple SSA
281value instead of a memory location.</p>
282
283</div>
284
285<!-- _______________________________________________________________________ -->
286<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
287
288<div class="doc_text">
289
290<p>It is important to note that this document describes 'well formed'
291LLVM assembly language. There is a difference between what the parser
292accepts and what is considered 'well formed'. For example, the
293following instruction is syntactically okay, but not well formed:</p>
294
295<div class="doc_code">
296<pre>
297%x = <a href="#i_add">add</a> i32 1, %x
298</pre>
299</div>
300
301<p>...because the definition of <tt>%x</tt> does not dominate all of
302its uses. The LLVM infrastructure provides a verification pass that may
303be used to verify that an LLVM module is well formed. This pass is
304automatically run by the parser after parsing input assembly and by
305the optimizer before it outputs bitcode. The violations pointed out
306by the verifier pass indicate bugs in transformation passes or input to
307the parser.</p>
308</div>
309
Chris Lattnera83fdc02007-10-03 17:34:29 +0000[diff] [blame]310<!-- Describe the typesetting conventions here. -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]311
312<!-- *********************************************************************** -->
313<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
314<!-- *********************************************************************** -->
315
316<div class="doc_text">
317
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]318 <p>LLVM identifiers come in two basic types: global and local. Global
319 identifiers (functions, global variables) begin with the @ character. Local
320 identifiers (register names, types) begin with the % character. Additionally,
321 there are three different formats for identifiers, for different purposes:
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]322
323<ol>
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]324 <li>Named values are represented as a string of characters with their prefix.
325 For example, %foo, @DivisionByZero, %a.really.long.identifier. The actual
326 regular expression used is '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]327 Identifiers which require other characters in their names can be surrounded
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]328 with quotes. In this way, anything except a <tt>&quot;</tt> character can
329 be used in a named value.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]330
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]331 <li>Unnamed values are represented as an unsigned numeric value with their
332 prefix. For example, %12, @2, %44.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]333
334 <li>Constants, which are described in a <a href="#constants">section about
335 constants</a>, below.</li>
336</ol>
337
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]338<p>LLVM requires that values start with a prefix for two reasons: Compilers
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]339don't need to worry about name clashes with reserved words, and the set of
340reserved words may be expanded in the future without penalty. Additionally,
341unnamed identifiers allow a compiler to quickly come up with a temporary
342variable without having to avoid symbol table conflicts.</p>
343
344<p>Reserved words in LLVM are very similar to reserved words in other
345languages. There are keywords for different opcodes
346('<tt><a href="#i_add">add</a></tt>',
347 '<tt><a href="#i_bitcast">bitcast</a></tt>',
348 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
349href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
350and others. These reserved words cannot conflict with variable names, because
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]351none of them start with a prefix character ('%' or '@').</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]352
353<p>Here is an example of LLVM code to multiply the integer variable
354'<tt>%X</tt>' by 8:</p>
355
356<p>The easy way:</p>
357
358<div class="doc_code">
359<pre>
360%result = <a href="#i_mul">mul</a> i32 %X, 8
361</pre>
362</div>
363
364<p>After strength reduction:</p>
365
366<div class="doc_code">
367<pre>
368%result = <a href="#i_shl">shl</a> i32 %X, i8 3
369</pre>
370</div>
371
372<p>And the hard way:</p>
373
374<div class="doc_code">
375<pre>
376<a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
377<a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
378%result = <a href="#i_add">add</a> i32 %1, %1
379</pre>
380</div>
381
382<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
383important lexical features of LLVM:</p>
384
385<ol>
386
387 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
388 line.</li>
389
390 <li>Unnamed temporaries are created when the result of a computation is not
391 assigned to a named value.</li>
392
393 <li>Unnamed temporaries are numbered sequentially</li>
394
395</ol>
396
397<p>...and it also shows a convention that we follow in this document. When
398demonstrating instructions, we will follow an instruction with a comment that
399defines the type and name of value produced. Comments are shown in italic
400text.</p>
401
402</div>
403
404<!-- *********************************************************************** -->
405<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
406<!-- *********************************************************************** -->
407
408<!-- ======================================================================= -->
409<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
410</div>
411
412<div class="doc_text">
413
414<p>LLVM programs are composed of "Module"s, each of which is a
415translation unit of the input programs. Each module consists of
416functions, global variables, and symbol table entries. Modules may be
417combined together with the LLVM linker, which merges function (and
418global variable) definitions, resolves forward declarations, and merges
419symbol table entries. Here is an example of the "hello world" module:</p>
420
421<div class="doc_code">
422<pre><i>; Declare the string constant as a global constant...</i>
423<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a
424 href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>
425
426<i>; External declaration of the puts function</i>
427<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
428
429<i>; Definition of main function</i>
430define i32 @main() { <i>; i32()* </i>
431 <i>; Convert [13x i8 ]* to i8 *...</i>
432 %cast210 = <a
433 href="#i_getelementptr">getelementptr</a> [13 x i8 ]* @.LC0, i64 0, i64 0 <i>; i8 *</i>
434
435 <i>; Call puts function to write out the string to stdout...</i>
436 <a
437 href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
438 <a
439 href="#i_ret">ret</a> i32 0<br>}<br>
440</pre>
441</div>
442
443<p>This example is made up of a <a href="#globalvars">global variable</a>
444named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
445function, and a <a href="#functionstructure">function definition</a>
446for "<tt>main</tt>".</p>
447
448<p>In general, a module is made up of a list of global values,
449where both functions and global variables are global values. Global values are
450represented by a pointer to a memory location (in this case, a pointer to an
451array of char, and a pointer to a function), and have one of the following <a
452href="#linkage">linkage types</a>.</p>
453
454</div>
455
456<!-- ======================================================================= -->
457<div class="doc_subsection">
458 <a name="linkage">Linkage Types</a>
459</div>
460
461<div class="doc_text">
462
463<p>
464All Global Variables and Functions have one of the following types of linkage:
465</p>
466
467<dl>
468
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame^]469 <dt><tt><b><a name="linkage_internal">internal</a></b></tt>: </dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]470
471 <dd>Global values with internal linkage are only directly accessible by
472 objects in the current module. In particular, linking code into a module with
473 an internal global value may cause the internal to be renamed as necessary to
474 avoid collisions. Because the symbol is internal to the module, all
475 references can be updated. This corresponds to the notion of the
476 '<tt>static</tt>' keyword in C.
477 </dd>
478
479 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
480
481 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
482 the same name when linkage occurs. This is typically used to implement
483 inline functions, templates, or other code which must be generated in each
484 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
485 allowed to be discarded.
486 </dd>
487
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame^]488 <dt><tt><b><a name="linkage_common">common</a></b></tt>: </dt>
489
490 <dd>"<tt>common</tt>" linkage is exactly the same as <tt>linkonce</tt>
491 linkage, except that unreferenced <tt>common</tt> globals may not be
492 discarded. This is used for globals that may be emitted in multiple
493 translation units, but that are not guaranteed to be emitted into every
494 translation unit that uses them. One example of this is tentative
495 definitions in C, such as "<tt>int X;</tt>" at global scope.
496 </dd>
497
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]498 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
499
Dale Johannesen96e7e092008-05-23 23:13:41 +0000[diff] [blame^]500 <dd>"<tt>weak</tt>" linkage is the same as <tt>common</tt> linkage, except
501 that some targets may choose to emit different assembly sequences for them
502 for target-dependent reasons. This is used for globals that are declared
503 "weak" in C source code.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]504 </dd>
505
506 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
507
508 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
509 pointer to array type. When two global variables with appending linkage are
510 linked together, the two global arrays are appended together. This is the
511 LLVM, typesafe, equivalent of having the system linker append together
512 "sections" with identical names when .o files are linked.
513 </dd>
514
515 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
516 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
517 until linked, if not linked, the symbol becomes null instead of being an
518 undefined reference.
519 </dd>
520
521 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
522
523 <dd>If none of the above identifiers are used, the global is externally
524 visible, meaning that it participates in linkage and can be used to resolve
525 external symbol references.
526 </dd>
527</dl>
528
529 <p>
530 The next two types of linkage are targeted for Microsoft Windows platform
531 only. They are designed to support importing (exporting) symbols from (to)
532 DLLs.
533 </p>
534
535 <dl>
536 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
537
538 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
539 or variable via a global pointer to a pointer that is set up by the DLL
540 exporting the symbol. On Microsoft Windows targets, the pointer name is
541 formed by combining <code>_imp__</code> and the function or variable name.
542 </dd>
543
544 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
545
546 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
547 pointer to a pointer in a DLL, so that it can be referenced with the
548 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
549 name is formed by combining <code>_imp__</code> and the function or variable
550 name.
551 </dd>
552
553</dl>
554
555<p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
556variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
557variable and was linked with this one, one of the two would be renamed,
558preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
559external (i.e., lacking any linkage declarations), they are accessible
560outside of the current module.</p>
561<p>It is illegal for a function <i>declaration</i>
562to have any linkage type other than "externally visible", <tt>dllimport</tt>,
563or <tt>extern_weak</tt>.</p>
564<p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
565linkages.
566</div>
567
568<!-- ======================================================================= -->
569<div class="doc_subsection">
570 <a name="callingconv">Calling Conventions</a>
571</div>
572
573<div class="doc_text">
574
575<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
576and <a href="#i_invoke">invokes</a> can all have an optional calling convention
577specified for the call. The calling convention of any pair of dynamic
578caller/callee must match, or the behavior of the program is undefined. The
579following calling conventions are supported by LLVM, and more may be added in
580the future:</p>
581
582<dl>
583 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
584
585 <dd>This calling convention (the default if no other calling convention is
586 specified) matches the target C calling conventions. This calling convention
587 supports varargs function calls and tolerates some mismatch in the declared
588 prototype and implemented declaration of the function (as does normal C).
589 </dd>
590
591 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
592
593 <dd>This calling convention attempts to make calls as fast as possible
594 (e.g. by passing things in registers). This calling convention allows the
595 target to use whatever tricks it wants to produce fast code for the target,
596 without having to conform to an externally specified ABI. Implementations of
Arnold Schwaighofer07444922008-05-14 09:17:12 +0000[diff] [blame]597 this convention should allow arbitrary
598 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> to be
599 supported. This calling convention does not support varargs and requires the
600 prototype of all callees to exactly match the prototype of the function
601 definition.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]602 </dd>
603
604 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
605
606 <dd>This calling convention attempts to make code in the caller as efficient
607 as possible under the assumption that the call is not commonly executed. As
608 such, these calls often preserve all registers so that the call does not break
609 any live ranges in the caller side. This calling convention does not support
610 varargs and requires the prototype of all callees to exactly match the
611 prototype of the function definition.
612 </dd>
613
614 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
615
616 <dd>Any calling convention may be specified by number, allowing
617 target-specific calling conventions to be used. Target specific calling
618 conventions start at 64.
619 </dd>
620</dl>
621
622<p>More calling conventions can be added/defined on an as-needed basis, to
623support pascal conventions or any other well-known target-independent
624convention.</p>
625
626</div>
627
628<!-- ======================================================================= -->
629<div class="doc_subsection">
630 <a name="visibility">Visibility Styles</a>
631</div>
632
633<div class="doc_text">
634
635<p>
636All Global Variables and Functions have one of the following visibility styles:
637</p>
638
639<dl>
640 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
641
642 <dd>On ELF, default visibility means that the declaration is visible to other
643 modules and, in shared libraries, means that the declared entity may be
644 overridden. On Darwin, default visibility means that the declaration is
645 visible to other modules. Default visibility corresponds to "external
646 linkage" in the language.
647 </dd>
648
649 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
650
651 <dd>Two declarations of an object with hidden visibility refer to the same
652 object if they are in the same shared object. Usually, hidden visibility
653 indicates that the symbol will not be placed into the dynamic symbol table,
654 so no other module (executable or shared library) can reference it
655 directly.
656 </dd>
657
658 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
659
660 <dd>On ELF, protected visibility indicates that the symbol will be placed in
661 the dynamic symbol table, but that references within the defining module will
662 bind to the local symbol. That is, the symbol cannot be overridden by another
663 module.
664 </dd>
665</dl>
666
667</div>
668
669<!-- ======================================================================= -->
670<div class="doc_subsection">
671 <a name="globalvars">Global Variables</a>
672</div>
673
674<div class="doc_text">
675
676<p>Global variables define regions of memory allocated at compilation time
677instead of run-time. Global variables may optionally be initialized, may have
678an explicit section to be placed in, and may have an optional explicit alignment
679specified. A variable may be defined as "thread_local", which means that it
680will not be shared by threads (each thread will have a separated copy of the
681variable). A variable may be defined as a global "constant," which indicates
682that the contents of the variable will <b>never</b> be modified (enabling better
683optimization, allowing the global data to be placed in the read-only section of
684an executable, etc). Note that variables that need runtime initialization
685cannot be marked "constant" as there is a store to the variable.</p>
686
687<p>
688LLVM explicitly allows <em>declarations</em> of global variables to be marked
689constant, even if the final definition of the global is not. This capability
690can be used to enable slightly better optimization of the program, but requires
691the language definition to guarantee that optimizations based on the
692'constantness' are valid for the translation units that do not include the
693definition.
694</p>
695
696<p>As SSA values, global variables define pointer values that are in
697scope (i.e. they dominate) all basic blocks in the program. Global
698variables always define a pointer to their "content" type because they
699describe a region of memory, and all memory objects in LLVM are
700accessed through pointers.</p>
701
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]702<p>A global variable may be declared to reside in a target-specifc numbered
703address space. For targets that support them, address spaces may affect how
704optimizations are performed and/or what target instructions are used to access
Christopher Lamb20a39e92007-12-12 08:44:39 +0000[diff] [blame]705the variable. The default address space is zero. The address space qualifier
706must precede any other attributes.</p>
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]707
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]708<p>LLVM allows an explicit section to be specified for globals. If the target
709supports it, it will emit globals to the section specified.</p>
710
711<p>An explicit alignment may be specified for a global. If not present, or if
712the alignment is set to zero, the alignment of the global is set by the target
713to whatever it feels convenient. If an explicit alignment is specified, the
714global is forced to have at least that much alignment. All alignments must be
715a power of 2.</p>
716
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]717<p>For example, the following defines a global in a numbered address space with
718an initializer, section, and alignment:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]719
720<div class="doc_code">
721<pre>
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]722@G = constant float 1.0 addrspace(5), section "foo", align 4
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]723</pre>
724</div>
725
726</div>
727
728
729<!-- ======================================================================= -->
730<div class="doc_subsection">
731 <a name="functionstructure">Functions</a>
732</div>
733
734<div class="doc_text">
735
736<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
737an optional <a href="#linkage">linkage type</a>, an optional
738<a href="#visibility">visibility style</a>, an optional
739<a href="#callingconv">calling convention</a>, a return type, an optional
740<a href="#paramattrs">parameter attribute</a> for the return type, a function
741name, a (possibly empty) argument list (each with optional
742<a href="#paramattrs">parameter attributes</a>), an optional section, an
Gordon Henriksend606f5b2007-12-10 03:30:21 +0000[diff] [blame]743optional alignment, an optional <a href="#gc">garbage collector name</a>, an
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]744opening curly brace, a list of basic blocks, and a closing curly brace.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]745
746LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
747optional <a href="#linkage">linkage type</a>, an optional
748<a href="#visibility">visibility style</a>, an optional
749<a href="#callingconv">calling convention</a>, a return type, an optional
750<a href="#paramattrs">parameter attribute</a> for the return type, a function
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]751name, a possibly empty list of arguments, an optional alignment, and an optional
Gordon Henriksend606f5b2007-12-10 03:30:21 +0000[diff] [blame]752<a href="#gc">garbage collector name</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]753
754<p>A function definition contains a list of basic blocks, forming the CFG for
755the function. Each basic block may optionally start with a label (giving the
756basic block a symbol table entry), contains a list of instructions, and ends
757with a <a href="#terminators">terminator</a> instruction (such as a branch or
758function return).</p>
759
760<p>The first basic block in a function is special in two ways: it is immediately
761executed on entrance to the function, and it is not allowed to have predecessor
762basic blocks (i.e. there can not be any branches to the entry block of a
763function). Because the block can have no predecessors, it also cannot have any
764<a href="#i_phi">PHI nodes</a>.</p>
765
766<p>LLVM allows an explicit section to be specified for functions. If the target
767supports it, it will emit functions to the section specified.</p>
768
769<p>An explicit alignment may be specified for a function. If not present, or if
770the alignment is set to zero, the alignment of the function is set by the target
771to whatever it feels convenient. If an explicit alignment is specified, the
772function is forced to have at least that much alignment. All alignments must be
773a power of 2.</p>
774
775</div>
776
777
778<!-- ======================================================================= -->
779<div class="doc_subsection">
780 <a name="aliasstructure">Aliases</a>
781</div>
782<div class="doc_text">
783 <p>Aliases act as "second name" for the aliasee value (which can be either
Anton Korobeynikov96822812008-03-22 08:36:14 +0000[diff] [blame]784 function, global variable, another alias or bitcast of global value). Aliases
785 may have an optional <a href="#linkage">linkage type</a>, and an
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]786 optional <a href="#visibility">visibility style</a>.</p>
787
788 <h5>Syntax:</h5>
789
790<div class="doc_code">
791<pre>
792@&lt;Name&gt; = [Linkage] [Visibility] alias &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
793</pre>
794</div>
795
796</div>
797
798
799
800<!-- ======================================================================= -->
801<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
802<div class="doc_text">
803 <p>The return type and each parameter of a function type may have a set of
804 <i>parameter attributes</i> associated with them. Parameter attributes are
805 used to communicate additional information about the result or parameters of
Duncan Sandsf5588dc2007-11-27 13:23:08 +0000[diff] [blame]806 a function. Parameter attributes are considered to be part of the function,
807 not of the function type, so functions with different parameter attributes
808 can have the same function type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]809
810 <p>Parameter attributes are simple keywords that follow the type specified. If
811 multiple parameter attributes are needed, they are space separated. For
812 example:</p>
813
814<div class="doc_code">
815<pre>
Duncan Sandsf5588dc2007-11-27 13:23:08 +0000[diff] [blame]816declare i32 @printf(i8* noalias , ...) nounwind
817declare i32 @atoi(i8*) nounwind readonly
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]818</pre>
819</div>
820
Duncan Sandsf5588dc2007-11-27 13:23:08 +0000[diff] [blame]821 <p>Note that any attributes for the function result (<tt>nounwind</tt>,
822 <tt>readonly</tt>) come immediately after the argument list.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]823
824 <p>Currently, only the following parameter attributes are defined:</p>
825 <dl>
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]826 <dt><tt>zeroext</tt></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]827 <dd>This indicates that the parameter should be zero extended just before
828 a call to this function.</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]829
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]830 <dt><tt>signext</tt></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]831 <dd>This indicates that the parameter should be sign extended just before
832 a call to this function.</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]833
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]834 <dt><tt>inreg</tt></dt>
835 <dd>This indicates that the parameter should be placed in register (if
836 possible) during assembling function call. Support for this attribute is
837 target-specific</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]838
839 <dt><tt>byval</tt></dt>
Chris Lattner04c86182008-01-15 04:34:22 +0000[diff] [blame]840 <dd>This indicates that the pointer parameter should really be passed by
841 value to the function. The attribute implies that a hidden copy of the
842 pointee is made between the caller and the callee, so the callee is unable
843 to modify the value in the callee. This attribute is only valid on llvm
844 pointer arguments. It is generally used to pass structs and arrays by
845 value, but is also valid on scalars (even though this is silly).</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]846
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]847 <dt><tt>sret</tt></dt>
Duncan Sands616cc032008-02-18 04:19:38 +0000[diff] [blame]848 <dd>This indicates that the pointer parameter specifies the address of a
849 structure that is the return value of the function in the source program.
Duncan Sandsef0e9e42008-03-17 12:17:41 +0000[diff] [blame]850 Loads and stores to the structure are assumed not to trap.
Duncan Sands616cc032008-02-18 04:19:38 +0000[diff] [blame]851 May only be applied to the first parameter.</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]852
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]853 <dt><tt>noalias</tt></dt>
Owen Andersonc4fc4cd2008-02-18 04:09:01 +0000[diff] [blame]854 <dd>This indicates that the parameter does not alias any global or any other
855 parameter. The caller is responsible for ensuring that this is the case,
856 usually by placing the value in a stack allocation.</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]857
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]858 <dt><tt>noreturn</tt></dt>
859 <dd>This function attribute indicates that the function never returns. This
860 indicates to LLVM that every call to this function should be treated as if
861 an <tt>unreachable</tt> instruction immediately followed the call.</dd>
Chris Lattner275e6be2008-01-11 06:20:47 +0000[diff] [blame]862
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]863 <dt><tt>nounwind</tt></dt>
Duncan Sandsef0e9e42008-03-17 12:17:41 +0000[diff] [blame]864 <dd>This function attribute indicates that no exceptions unwind out of the
865 function. Usually this is because the function makes no use of exceptions,
866 but it may also be that the function catches any exceptions thrown when
867 executing it.</dd>
868
Duncan Sands4ee46812007-07-27 19:57:41 +0000[diff] [blame]869 <dt><tt>nest</tt></dt>
870 <dd>This indicates that the parameter can be excised using the
871 <a href="#int_trampoline">trampoline intrinsics</a>.</dd>
Duncan Sands13e13f82007-11-22 20:23:04 +0000[diff] [blame]872 <dt><tt>readonly</tt></dt>
Duncan Sandsd69c0e62007-11-14 21:14:02 +0000[diff] [blame]873 <dd>This function attribute indicates that the function has no side-effects
Duncan Sands13e13f82007-11-22 20:23:04 +0000[diff] [blame]874 except for producing a return value or throwing an exception. The value
875 returned must only depend on the function arguments and/or global variables.
876 It may use values obtained by dereferencing pointers.</dd>
877 <dt><tt>readnone</tt></dt>
878 <dd>A <tt>readnone</tt> function has the same restrictions as a <tt>readonly</tt>
Duncan Sandsd69c0e62007-11-14 21:14:02 +0000[diff] [blame]879 function, but in addition it is not allowed to dereference any pointer arguments
880 or global variables.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]881 </dl>
882
883</div>
884
885<!-- ======================================================================= -->
886<div class="doc_subsection">
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]887 <a name="gc">Garbage Collector Names</a>
888</div>
889
890<div class="doc_text">
891<p>Each function may specify a garbage collector name, which is simply a
892string.</p>
893
894<div class="doc_code"><pre
895>define void @f() gc "name" { ...</pre></div>
896
897<p>The compiler declares the supported values of <i>name</i>. Specifying a
898collector which will cause the compiler to alter its output in order to support
899the named garbage collection algorithm.</p>
900</div>
901
902<!-- ======================================================================= -->
903<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]904 <a name="moduleasm">Module-Level Inline Assembly</a>
905</div>
906
907<div class="doc_text">
908<p>
909Modules may contain "module-level inline asm" blocks, which corresponds to the
910GCC "file scope inline asm" blocks. These blocks are internally concatenated by
911LLVM and treated as a single unit, but may be separated in the .ll file if
912desired. The syntax is very simple:
913</p>
914
915<div class="doc_code">
916<pre>
917module asm "inline asm code goes here"
918module asm "more can go here"
919</pre>
920</div>
921
922<p>The strings can contain any character by escaping non-printable characters.
923 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
924 for the number.
925</p>
926
927<p>
928 The inline asm code is simply printed to the machine code .s file when
929 assembly code is generated.
930</p>
931</div>
932
933<!-- ======================================================================= -->
934<div class="doc_subsection">
935 <a name="datalayout">Data Layout</a>
936</div>
937
938<div class="doc_text">
939<p>A module may specify a target specific data layout string that specifies how
940data is to be laid out in memory. The syntax for the data layout is simply:</p>
941<pre> target datalayout = "<i>layout specification</i>"</pre>
942<p>The <i>layout specification</i> consists of a list of specifications
943separated by the minus sign character ('-'). Each specification starts with a
944letter and may include other information after the letter to define some
945aspect of the data layout. The specifications accepted are as follows: </p>
946<dl>
947 <dt><tt>E</tt></dt>
948 <dd>Specifies that the target lays out data in big-endian form. That is, the
949 bits with the most significance have the lowest address location.</dd>
950 <dt><tt>e</tt></dt>
951 <dd>Specifies that hte target lays out data in little-endian form. That is,
952 the bits with the least significance have the lowest address location.</dd>
953 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
954 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
955 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
956 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
957 too.</dd>
958 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
959 <dd>This specifies the alignment for an integer type of a given bit
960 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
961 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
962 <dd>This specifies the alignment for a vector type of a given bit
963 <i>size</i>.</dd>
964 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
965 <dd>This specifies the alignment for a floating point type of a given bit
966 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
967 (double).</dd>
968 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
969 <dd>This specifies the alignment for an aggregate type of a given bit
970 <i>size</i>.</dd>
971</dl>
972<p>When constructing the data layout for a given target, LLVM starts with a
973default set of specifications which are then (possibly) overriden by the
974specifications in the <tt>datalayout</tt> keyword. The default specifications
975are given in this list:</p>
976<ul>
977 <li><tt>E</tt> - big endian</li>
978 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
979 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
980 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
981 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
982 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
983 <li><tt>i64:32:64</tt> - i64 has abi alignment of 32-bits but preferred
984 alignment of 64-bits</li>
985 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
986 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
987 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
988 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
989 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
990</ul>
991<p>When llvm is determining the alignment for a given type, it uses the
992following rules:
993<ol>
994 <li>If the type sought is an exact match for one of the specifications, that
995 specification is used.</li>
996 <li>If no match is found, and the type sought is an integer type, then the
997 smallest integer type that is larger than the bitwidth of the sought type is
998 used. If none of the specifications are larger than the bitwidth then the the
999 largest integer type is used. For example, given the default specifications
1000 above, the i7 type will use the alignment of i8 (next largest) while both
1001 i65 and i256 will use the alignment of i64 (largest specified).</li>
1002 <li>If no match is found, and the type sought is a vector type, then the
1003 largest vector type that is smaller than the sought vector type will be used
1004 as a fall back. This happens because <128 x double> can be implemented in
1005 terms of 64 <2 x double>, for example.</li>
1006</ol>
1007</div>
1008
1009<!-- *********************************************************************** -->
1010<div class="doc_section"> <a name="typesystem">Type System</a> </div>
1011<!-- *********************************************************************** -->
1012
1013<div class="doc_text">
1014
1015<p>The LLVM type system is one of the most important features of the
1016intermediate representation. Being typed enables a number of
1017optimizations to be performed on the IR directly, without having to do
1018extra analyses on the side before the transformation. A strong type
1019system makes it easier to read the generated code and enables novel
1020analyses and transformations that are not feasible to perform on normal
1021three address code representations.</p>
1022
1023</div>
1024
1025<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1026<div class="doc_subsection"> <a name="t_classifications">Type
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1027Classifications</a> </div>
1028<div class="doc_text">
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1029<p>The types fall into a few useful
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1030classifications:</p>
1031
1032<table border="1" cellspacing="0" cellpadding="4">
1033 <tbody>
1034 <tr><th>Classification</th><th>Types</th></tr>
1035 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1036 <td><a href="#t_integer">integer</a></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1037 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
1038 </tr>
1039 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1040 <td><a href="#t_floating">floating point</a></td>
1041 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1042 </tr>
1043 <tr>
1044 <td><a name="t_firstclass">first class</a></td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1045 <td><a href="#t_integer">integer</a>,
1046 <a href="#t_floating">floating point</a>,
1047 <a href="#t_pointer">pointer</a>,
1048 <a href="#t_vector">vector</a>
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]1049 <a href="#t_struct">structure</a>,
1050 <a href="#t_array">array</a>,
Dan Gohmand13951e2008-05-23 22:50:26 +0000[diff] [blame]1051 <a href="#t_label">label</a>.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1052 </td>
1053 </tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1054 <tr>
1055 <td><a href="#t_primitive">primitive</a></td>
1056 <td><a href="#t_label">label</a>,
1057 <a href="#t_void">void</a>,
1058 <a href="#t_integer">integer</a>,
1059 <a href="#t_floating">floating point</a>.</td>
1060 </tr>
1061 <tr>
1062 <td><a href="#t_derived">derived</a></td>
1063 <td><a href="#t_integer">integer</a>,
1064 <a href="#t_array">array</a>,
1065 <a href="#t_function">function</a>,
1066 <a href="#t_pointer">pointer</a>,
1067 <a href="#t_struct">structure</a>,
1068 <a href="#t_pstruct">packed structure</a>,
1069 <a href="#t_vector">vector</a>,
1070 <a href="#t_opaque">opaque</a>.
1071 </tr>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1072 </tbody>
1073</table>
1074
1075<p>The <a href="#t_firstclass">first class</a> types are perhaps the
1076most important. Values of these types are the only ones which can be
1077produced by instructions, passed as arguments, or used as operands to
Dan Gohman4f29e422008-05-23 21:53:15 +0000[diff] [blame]1078instructions.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1079</div>
1080
1081<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1082<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame]1083
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1084<div class="doc_text">
1085<p>The primitive types are the fundamental building blocks of the LLVM
1086system.</p>
1087
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame]1088</div>
1089
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1090<!-- _______________________________________________________________________ -->
1091<div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1092
1093<div class="doc_text">
1094 <table>
1095 <tbody>
1096 <tr><th>Type</th><th>Description</th></tr>
1097 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1098 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1099 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1100 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1101 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1102 </tbody>
1103 </table>
1104</div>
1105
1106<!-- _______________________________________________________________________ -->
1107<div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1108
1109<div class="doc_text">
1110<h5>Overview:</h5>
1111<p>The void type does not represent any value and has no size.</p>
1112
1113<h5>Syntax:</h5>
1114
1115<pre>
1116 void
1117</pre>
1118</div>
1119
1120<!-- _______________________________________________________________________ -->
1121<div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1122
1123<div class="doc_text">
1124<h5>Overview:</h5>
1125<p>The label type represents code labels.</p>
1126
1127<h5>Syntax:</h5>
1128
1129<pre>
1130 label
1131</pre>
1132</div>
1133
1134
1135<!-- ======================================================================= -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1136<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
1137
1138<div class="doc_text">
1139
1140<p>The real power in LLVM comes from the derived types in the system.
1141This is what allows a programmer to represent arrays, functions,
1142pointers, and other useful types. Note that these derived types may be
1143recursive: For example, it is possible to have a two dimensional array.</p>
1144
1145</div>
1146
1147<!-- _______________________________________________________________________ -->
1148<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1149
1150<div class="doc_text">
1151
1152<h5>Overview:</h5>
1153<p>The integer type is a very simple derived type that simply specifies an
1154arbitrary bit width for the integer type desired. Any bit width from 1 bit to
11552^23-1 (about 8 million) can be specified.</p>
1156
1157<h5>Syntax:</h5>
1158
1159<pre>
1160 iN
1161</pre>
1162
1163<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1164value.</p>
1165
1166<h5>Examples:</h5>
1167<table class="layout">
Chris Lattner251ab812007-12-18 06:18:21 +0000[diff] [blame]1168 <tbody>
1169 <tr>
1170 <td><tt>i1</tt></td>
1171 <td>a single-bit integer.</td>
1172 </tr><tr>
1173 <td><tt>i32</tt></td>
1174 <td>a 32-bit integer.</td>
1175 </tr><tr>
1176 <td><tt>i1942652</tt></td>
1177 <td>a really big integer of over 1 million bits.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1178 </tr>
Chris Lattner251ab812007-12-18 06:18:21 +0000[diff] [blame]1179 </tbody>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1180</table>
1181</div>
1182
1183<!-- _______________________________________________________________________ -->
1184<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
1185
1186<div class="doc_text">
1187
1188<h5>Overview:</h5>
1189
1190<p>The array type is a very simple derived type that arranges elements
1191sequentially in memory. The array type requires a size (number of
1192elements) and an underlying data type.</p>
1193
1194<h5>Syntax:</h5>
1195
1196<pre>
1197 [&lt;# elements&gt; x &lt;elementtype&gt;]
1198</pre>
1199
1200<p>The number of elements is a constant integer value; elementtype may
1201be any type with a size.</p>
1202
1203<h5>Examples:</h5>
1204<table class="layout">
1205 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1206 <td class="left"><tt>[40 x i32]</tt></td>
1207 <td class="left">Array of 40 32-bit integer values.</td>
1208 </tr>
1209 <tr class="layout">
1210 <td class="left"><tt>[41 x i32]</tt></td>
1211 <td class="left">Array of 41 32-bit integer values.</td>
1212 </tr>
1213 <tr class="layout">
1214 <td class="left"><tt>[4 x i8]</tt></td>
1215 <td class="left">Array of 4 8-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1216 </tr>
1217</table>
1218<p>Here are some examples of multidimensional arrays:</p>
1219<table class="layout">
1220 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1221 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1222 <td class="left">3x4 array of 32-bit integer values.</td>
1223 </tr>
1224 <tr class="layout">
1225 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1226 <td class="left">12x10 array of single precision floating point values.</td>
1227 </tr>
1228 <tr class="layout">
1229 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1230 <td class="left">2x3x4 array of 16-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1231 </tr>
1232</table>
1233
1234<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1235length array. Normally, accesses past the end of an array are undefined in
1236LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1237As a special case, however, zero length arrays are recognized to be variable
1238length. This allows implementation of 'pascal style arrays' with the LLVM
1239type "{ i32, [0 x float]}", for example.</p>
1240
1241</div>
1242
1243<!-- _______________________________________________________________________ -->
1244<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
1245<div class="doc_text">
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1246
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1247<h5>Overview:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1248
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1249<p>The function type can be thought of as a function signature. It
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1250consists of a return type and a list of formal parameter types. The
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1251return 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]1252If 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]1253class types, and the struct must have at least one element.</p>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1254
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1255<h5>Syntax:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1256
1257<pre>
1258 &lt;returntype list&gt; (&lt;parameter list&gt;)
1259</pre>
1260
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1261<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
1262specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1263which indicates that the function takes a variable number of arguments.
1264Variable argument functions can access their arguments with the <a
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1265 href="#int_varargs">variable argument handling intrinsic</a> functions.
1266'<tt>&lt;returntype list&gt;</tt>' is a comma-separated list of
1267<a href="#t_firstclass">first class</a> type specifiers.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1268
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1269<h5>Examples:</h5>
1270<table class="layout">
1271 <tr class="layout">
1272 <td class="left"><tt>i32 (i32)</tt></td>
1273 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
1274 </td>
1275 </tr><tr class="layout">
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]1276 <td class="left"><tt>float&nbsp;(i16&nbsp;signext,&nbsp;i32&nbsp;*)&nbsp;*
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1277 </tt></td>
1278 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1279 an <tt>i16</tt> that should be sign extended and a
1280 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
1281 <tt>float</tt>.
1282 </td>
1283 </tr><tr class="layout">
1284 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1285 <td class="left">A vararg function that takes at least one
1286 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1287 which returns an integer. This is the signature for <tt>printf</tt> in
1288 LLVM.
1289 </td>
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1290 </tr><tr class="layout">
1291 <td class="left"><tt>{i32, i32} (i32)</tt></td>
Devang Patel6dddba22008-03-24 18:10:52 +0000[diff] [blame]1292 <td class="left">A function taking an <tt>i32></tt>, returning two
1293 <tt> i32 </tt> values as an aggregate of type <tt>{ i32, i32 }</tt>
Devang Pateld4ba41d2008-03-24 05:35:41 +0000[diff] [blame]1294 </td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1295 </tr>
1296</table>
1297
1298</div>
1299<!-- _______________________________________________________________________ -->
1300<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
1301<div class="doc_text">
1302<h5>Overview:</h5>
1303<p>The structure type is used to represent a collection of data members
1304together in memory. The packing of the field types is defined to match
1305the ABI of the underlying processor. The elements of a structure may
1306be any type that has a size.</p>
1307<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1308and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1309field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1310instruction.</p>
1311<h5>Syntax:</h5>
1312<pre> { &lt;type list&gt; }<br></pre>
1313<h5>Examples:</h5>
1314<table class="layout">
1315 <tr class="layout">
1316 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1317 <td class="left">A triple of three <tt>i32</tt> values</td>
1318 </tr><tr class="layout">
1319 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1320 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1321 second element is a <a href="#t_pointer">pointer</a> to a
1322 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1323 an <tt>i32</tt>.</td>
1324 </tr>
1325</table>
1326</div>
1327
1328<!-- _______________________________________________________________________ -->
1329<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1330</div>
1331<div class="doc_text">
1332<h5>Overview:</h5>
1333<p>The packed structure type is used to represent a collection of data members
1334together in memory. There is no padding between fields. Further, the alignment
1335of a packed structure is 1 byte. The elements of a packed structure may
1336be any type that has a size.</p>
1337<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1338and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1339field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1340instruction.</p>
1341<h5>Syntax:</h5>
1342<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1343<h5>Examples:</h5>
1344<table class="layout">
1345 <tr class="layout">
1346 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1347 <td class="left">A triple of three <tt>i32</tt> values</td>
1348 </tr><tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1349 <td class="left"><tt>&lt; { float, i32 (i32)* } &gt;</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1350 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1351 second element is a <a href="#t_pointer">pointer</a> to a
1352 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1353 an <tt>i32</tt>.</td>
1354 </tr>
1355</table>
1356</div>
1357
1358<!-- _______________________________________________________________________ -->
1359<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
1360<div class="doc_text">
1361<h5>Overview:</h5>
1362<p>As in many languages, the pointer type represents a pointer or
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]1363reference to another object, which must live in memory. Pointer types may have
1364an optional address space attribute defining the target-specific numbered
1365address space where the pointed-to object resides. The default address space is
1366zero.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1367<h5>Syntax:</h5>
1368<pre> &lt;type&gt; *<br></pre>
1369<h5>Examples:</h5>
1370<table class="layout">
1371 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1372 <td class="left"><tt>[4x i32]*</tt></td>
1373 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1374 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1375 </tr>
1376 <tr class="layout">
1377 <td class="left"><tt>i32 (i32 *) *</tt></td>
1378 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1379 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1380 <tt>i32</tt>.</td>
1381 </tr>
1382 <tr class="layout">
1383 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1384 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1385 that resides in address space #5.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1386 </tr>
1387</table>
1388</div>
1389
1390<!-- _______________________________________________________________________ -->
1391<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
1392<div class="doc_text">
1393
1394<h5>Overview:</h5>
1395
1396<p>A vector type is a simple derived type that represents a vector
1397of elements. Vector types are used when multiple primitive data
1398are operated in parallel using a single instruction (SIMD).
1399A vector type requires a size (number of
1400elements) and an underlying primitive data type. Vectors must have a power
1401of two length (1, 2, 4, 8, 16 ...). Vector types are
1402considered <a href="#t_firstclass">first class</a>.</p>
1403
1404<h5>Syntax:</h5>
1405
1406<pre>
1407 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1408</pre>
1409
1410<p>The number of elements is a constant integer value; elementtype may
1411be any integer or floating point type.</p>
1412
1413<h5>Examples:</h5>
1414
1415<table class="layout">
1416 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1417 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1418 <td class="left">Vector of 4 32-bit integer values.</td>
1419 </tr>
1420 <tr class="layout">
1421 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1422 <td class="left">Vector of 8 32-bit floating-point values.</td>
1423 </tr>
1424 <tr class="layout">
1425 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1426 <td class="left">Vector of 2 64-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1427 </tr>
1428</table>
1429</div>
1430
1431<!-- _______________________________________________________________________ -->
1432<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1433<div class="doc_text">
1434
1435<h5>Overview:</h5>
1436
1437<p>Opaque types are used to represent unknown types in the system. This
Gordon Henriksenda0706e2007-10-14 00:34:53 +0000[diff] [blame]1438corresponds (for example) to the C notion of a forward declared structure type.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1439In LLVM, opaque types can eventually be resolved to any type (not just a
1440structure type).</p>
1441
1442<h5>Syntax:</h5>
1443
1444<pre>
1445 opaque
1446</pre>
1447
1448<h5>Examples:</h5>
1449
1450<table class="layout">
1451 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1452 <td class="left"><tt>opaque</tt></td>
1453 <td class="left">An opaque type.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1454 </tr>
1455</table>
1456</div>
1457
1458
1459<!-- *********************************************************************** -->
1460<div class="doc_section"> <a name="constants">Constants</a> </div>
1461<!-- *********************************************************************** -->
1462
1463<div class="doc_text">
1464
1465<p>LLVM has several different basic types of constants. This section describes
1466them all and their syntax.</p>
1467
1468</div>
1469
1470<!-- ======================================================================= -->
1471<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
1472
1473<div class="doc_text">
1474
1475<dl>
1476 <dt><b>Boolean constants</b></dt>
1477
1478 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
1479 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
1480 </dd>
1481
1482 <dt><b>Integer constants</b></dt>
1483
1484 <dd>Standard integers (such as '4') are constants of the <a
1485 href="#t_integer">integer</a> type. Negative numbers may be used with
1486 integer types.
1487 </dd>
1488
1489 <dt><b>Floating point constants</b></dt>
1490
1491 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1492 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]1493 notation (see below). The assembler requires the exact decimal value of
1494 a floating-point constant. For example, the assembler accepts 1.25 but
1495 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
1496 constants must have a <a href="#t_floating">floating point</a> type. </dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1497
1498 <dt><b>Null pointer constants</b></dt>
1499
1500 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
1501 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1502
1503</dl>
1504
1505<p>The one non-intuitive notation for constants is the optional hexadecimal form
1506of floating point constants. For example, the form '<tt>double
15070x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
15084.5e+15</tt>'. The only time hexadecimal floating point constants are required
1509(and the only time that they are generated by the disassembler) is when a
1510floating point constant must be emitted but it cannot be represented as a
1511decimal floating point number. For example, NaN's, infinities, and other
1512special values are represented in their IEEE hexadecimal format so that
1513assembly and disassembly do not cause any bits to change in the constants.</p>
1514
1515</div>
1516
1517<!-- ======================================================================= -->
1518<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1519</div>
1520
1521<div class="doc_text">
1522<p>Aggregate constants arise from aggregation of simple constants
1523and smaller aggregate constants.</p>
1524
1525<dl>
1526 <dt><b>Structure constants</b></dt>
1527
1528 <dd>Structure constants are represented with notation similar to structure
1529 type definitions (a comma separated list of elements, surrounded by braces
Chris Lattner6c8de962007-12-25 20:34:52 +0000[diff] [blame]1530 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
1531 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]1532 must have <a href="#t_struct">structure type</a>, and the number and
1533 types of elements must match those specified by the type.
1534 </dd>
1535
1536 <dt><b>Array constants</b></dt>
1537
1538 <dd>Array constants are represented with notation similar to array type
1539 definitions (a comma separated list of elements, surrounded by square brackets
1540 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
1541 constants must have <a href="#t_array">array type</a>, and the number and
1542 types of elements must match those specified by the type.
1543 </dd>
1544
1545 <dt><b>Vector constants</b></dt>
1546
1547 <dd>Vector constants are represented with notation similar to vector type
1548 definitions (a comma separated list of elements, surrounded by
1549 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
1550 i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must have <a
1551 href="#t_vector">vector type</a>, and the number and types of elements must
1552 match those specified by the type.
1553 </dd>
1554
1555 <dt><b>Zero initialization</b></dt>
1556
1557 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1558 value to zero of <em>any</em> type, including scalar and aggregate types.
1559 This is often used to avoid having to print large zero initializers (e.g. for
1560 large arrays) and is always exactly equivalent to using explicit zero
1561 initializers.
1562 </dd>
1563</dl>
1564
1565</div>
1566
1567<!-- ======================================================================= -->
1568<div class="doc_subsection">
1569 <a name="globalconstants">Global Variable and Function Addresses</a>
1570</div>
1571
1572<div class="doc_text">
1573
1574<p>The addresses of <a href="#globalvars">global variables</a> and <a
1575href="#functionstructure">functions</a> are always implicitly valid (link-time)
1576constants. These constants are explicitly referenced when the <a
1577href="#identifiers">identifier for the global</a> is used and always have <a
1578href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1579file:</p>
1580
1581<div class="doc_code">
1582<pre>
1583@X = global i32 17
1584@Y = global i32 42
1585@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
1586</pre>
1587</div>
1588
1589</div>
1590
1591<!-- ======================================================================= -->
1592<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
1593<div class="doc_text">
1594 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
1595 no specific value. Undefined values may be of any type and be used anywhere
1596 a constant is permitted.</p>
1597
1598 <p>Undefined values indicate to the compiler that the program is well defined
1599 no matter what value is used, giving the compiler more freedom to optimize.
1600 </p>
1601</div>
1602
1603<!-- ======================================================================= -->
1604<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1605</div>
1606
1607<div class="doc_text">
1608
1609<p>Constant expressions are used to allow expressions involving other constants
1610to be used as constants. Constant expressions may be of any <a
1611href="#t_firstclass">first class</a> type and may involve any LLVM operation
1612that does not have side effects (e.g. load and call are not supported). The
1613following is the syntax for constant expressions:</p>
1614
1615<dl>
1616 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1617 <dd>Truncate a constant to another type. The bit size of CST must be larger
1618 than the bit size of TYPE. Both types must be integers.</dd>
1619
1620 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1621 <dd>Zero extend a constant to another type. The bit size of CST must be
1622 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1623
1624 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1625 <dd>Sign extend a constant to another type. The bit size of CST must be
1626 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1627
1628 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1629 <dd>Truncate a floating point constant to another floating point type. The
1630 size of CST must be larger than the size of TYPE. Both types must be
1631 floating point.</dd>
1632
1633 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1634 <dd>Floating point extend a constant to another type. The size of CST must be
1635 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1636
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]1637 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1638 <dd>Convert a floating point constant to the corresponding unsigned integer
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1639 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1640 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1641 of the same number of elements. If the value won't fit in the integer type,
1642 the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1643
1644 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
1645 <dd>Convert a floating point constant to the corresponding signed integer
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1646 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1647 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1648 of the same number of elements. If the value won't fit in the integer type,
1649 the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1650
1651 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
1652 <dd>Convert an unsigned integer constant to the corresponding floating point
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1653 constant. TYPE must be a scalar or vector floating point type. CST must be of
1654 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1655 of the same number of elements. If the value won't fit in the floating point
1656 type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1657
1658 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
1659 <dd>Convert a signed integer constant to the corresponding floating point
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1660 constant. TYPE must be a scalar or vector floating point type. CST must be of
1661 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1662 of the same number of elements. If the value won't fit in the floating point
1663 type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1664
1665 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1666 <dd>Convert a pointer typed constant to the corresponding integer constant
1667 TYPE must be an integer type. CST must be of pointer type. The CST value is
1668 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1669
1670 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1671 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1672 pointer type. CST must be of integer type. The CST value is zero extended,
1673 truncated, or unchanged to make it fit in a pointer size. This one is
1674 <i>really</i> dangerous!</dd>
1675
1676 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
1677 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1678 identical (same number of bits). The conversion is done as if the CST value
1679 was stored to memory and read back as TYPE. In other words, no bits change
1680 with this operator, just the type. This can be used for conversion of
1681 vector types to any other type, as long as they have the same bit width. For
1682 pointers it is only valid to cast to another pointer type.
1683 </dd>
1684
1685 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1686
1687 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1688 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1689 instruction, the index list may have zero or more indexes, which are required
1690 to make sense for the type of "CSTPTR".</dd>
1691
1692 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1693
1694 <dd>Perform the <a href="#i_select">select operation</a> on
1695 constants.</dd>
1696
1697 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1698 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1699
1700 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1701 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
1702
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]1703 <dt><b><tt>vicmp COND ( VAL1, VAL2 )</tt></b></dt>
1704 <dd>Performs the <a href="#i_vicmp">vicmp operation</a> on constants.</dd>
1705
1706 <dt><b><tt>vfcmp COND ( VAL1, VAL2 )</tt></b></dt>
1707 <dd>Performs the <a href="#i_vfcmp">vfcmp operation</a> on constants.</dd>
1708
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1709 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1710
1711 <dd>Perform the <a href="#i_extractelement">extractelement
1712 operation</a> on constants.
1713
1714 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1715
1716 <dd>Perform the <a href="#i_insertelement">insertelement
1717 operation</a> on constants.</dd>
1718
1719
1720 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1721
1722 <dd>Perform the <a href="#i_shufflevector">shufflevector
1723 operation</a> on constants.</dd>
1724
1725 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1726
1727 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1728 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
1729 binary</a> operations. The constraints on operands are the same as those for
1730 the corresponding instruction (e.g. no bitwise operations on floating point
1731 values are allowed).</dd>
1732</dl>
1733</div>
1734
1735<!-- *********************************************************************** -->
1736<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1737<!-- *********************************************************************** -->
1738
1739<!-- ======================================================================= -->
1740<div class="doc_subsection">
1741<a name="inlineasm">Inline Assembler Expressions</a>
1742</div>
1743
1744<div class="doc_text">
1745
1746<p>
1747LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1748Module-Level Inline Assembly</a>) through the use of a special value. This
1749value represents the inline assembler as a string (containing the instructions
1750to emit), a list of operand constraints (stored as a string), and a flag that
1751indicates whether or not the inline asm expression has side effects. An example
1752inline assembler expression is:
1753</p>
1754
1755<div class="doc_code">
1756<pre>
1757i32 (i32) asm "bswap $0", "=r,r"
1758</pre>
1759</div>
1760
1761<p>
1762Inline assembler expressions may <b>only</b> be used as the callee operand of
1763a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1764</p>
1765
1766<div class="doc_code">
1767<pre>
1768%X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
1769</pre>
1770</div>
1771
1772<p>
1773Inline asms with side effects not visible in the constraint list must be marked
1774as having side effects. This is done through the use of the
1775'<tt>sideeffect</tt>' keyword, like so:
1776</p>
1777
1778<div class="doc_code">
1779<pre>
1780call void asm sideeffect "eieio", ""()
1781</pre>
1782</div>
1783
1784<p>TODO: The format of the asm and constraints string still need to be
1785documented here. Constraints on what can be done (e.g. duplication, moving, etc
1786need to be documented).
1787</p>
1788
1789</div>
1790
1791<!-- *********************************************************************** -->
1792<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1793<!-- *********************************************************************** -->
1794
1795<div class="doc_text">
1796
1797<p>The LLVM instruction set consists of several different
1798classifications of instructions: <a href="#terminators">terminator
1799instructions</a>, <a href="#binaryops">binary instructions</a>,
1800<a href="#bitwiseops">bitwise binary instructions</a>, <a
1801 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1802instructions</a>.</p>
1803
1804</div>
1805
1806<!-- ======================================================================= -->
1807<div class="doc_subsection"> <a name="terminators">Terminator
1808Instructions</a> </div>
1809
1810<div class="doc_text">
1811
1812<p>As mentioned <a href="#functionstructure">previously</a>, every
1813basic block in a program ends with a "Terminator" instruction, which
1814indicates which block should be executed after the current block is
1815finished. These terminator instructions typically yield a '<tt>void</tt>'
1816value: they produce control flow, not values (the one exception being
1817the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
1818<p>There are six different terminator instructions: the '<a
1819 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1820instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
1821the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1822 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1823 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
1824
1825</div>
1826
1827<!-- _______________________________________________________________________ -->
1828<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1829Instruction</a> </div>
1830<div class="doc_text">
1831<h5>Syntax:</h5>
1832<pre> ret &lt;type&gt; &lt;value&gt; <i>; Return a value from a non-void function</i>
1833 ret void <i>; Return from void function</i>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1834 ret &lt;type&gt; &lt;value&gt;, &lt;type&gt; &lt;value&gt; <i>; Return two values from a non-void function </i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1835</pre>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1836
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1837<h5>Overview:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1838
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1839<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
1840value) from a function back to the caller.</p>
1841<p>There are two forms of the '<tt>ret</tt>' instruction: one that
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1842returns value(s) and then causes control flow, and one that just causes
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1843control flow to occur.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1844
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1845<h5>Arguments:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1846
1847<p>The '<tt>ret</tt>' instruction may return zero, one or multiple values.
1848The type of each return value must be a '<a href="#t_firstclass">first
1849class</a>' type. Note that a function is not <a href="#wellformed">well
1850formed</a> if there exists a '<tt>ret</tt>' instruction inside of the
1851function that returns values that do not match the return type of the
1852function.</p>
1853
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1854<h5>Semantics:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1855
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1856<p>When the '<tt>ret</tt>' instruction is executed, control flow
1857returns back to the calling function's context. If the caller is a "<a
1858 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
1859the instruction after the call. If the caller was an "<a
1860 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
1861at the beginning of the "normal" destination block. If the instruction
1862returns a value, that value shall set the call or invoke instruction's
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1863return value. If the instruction returns multiple values then these
Devang Patelec8a5b02008-03-11 05:51:59 +0000[diff] [blame]1864values can only be accessed through a '<a href="#i_getresult"><tt>getresult</tt>
1865</a>' instruction.</p>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1866
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1867<h5>Example:</h5>
Chris Lattner43030e72008-04-23 04:59:35 +0000[diff] [blame]1868
1869<pre>
1870 ret i32 5 <i>; Return an integer value of 5</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1871 ret void <i>; Return from a void function</i>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1872 ret i32 4, i8 2 <i>; Return two values 4 and 2 </i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1873</pre>
1874</div>
1875<!-- _______________________________________________________________________ -->
1876<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
1877<div class="doc_text">
1878<h5>Syntax:</h5>
1879<pre> br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
1880</pre>
1881<h5>Overview:</h5>
1882<p>The '<tt>br</tt>' instruction is used to cause control flow to
1883transfer to a different basic block in the current function. There are
1884two forms of this instruction, corresponding to a conditional branch
1885and an unconditional branch.</p>
1886<h5>Arguments:</h5>
1887<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
1888single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
1889unconditional form of the '<tt>br</tt>' instruction takes a single
1890'<tt>label</tt>' value as a target.</p>
1891<h5>Semantics:</h5>
1892<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
1893argument is evaluated. If the value is <tt>true</tt>, control flows
1894to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1895control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
1896<h5>Example:</h5>
1897<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
1898 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
1899</div>
1900<!-- _______________________________________________________________________ -->
1901<div class="doc_subsubsection">
1902 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1903</div>
1904
1905<div class="doc_text">
1906<h5>Syntax:</h5>
1907
1908<pre>
1909 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1910</pre>
1911
1912<h5>Overview:</h5>
1913
1914<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1915several different places. It is a generalization of the '<tt>br</tt>'
1916instruction, allowing a branch to occur to one of many possible
1917destinations.</p>
1918
1919
1920<h5>Arguments:</h5>
1921
1922<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1923comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1924an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1925table is not allowed to contain duplicate constant entries.</p>
1926
1927<h5>Semantics:</h5>
1928
1929<p>The <tt>switch</tt> instruction specifies a table of values and
1930destinations. When the '<tt>switch</tt>' instruction is executed, this
1931table is searched for the given value. If the value is found, control flow is
1932transfered to the corresponding destination; otherwise, control flow is
1933transfered to the default destination.</p>
1934
1935<h5>Implementation:</h5>
1936
1937<p>Depending on properties of the target machine and the particular
1938<tt>switch</tt> instruction, this instruction may be code generated in different
1939ways. For example, it could be generated as a series of chained conditional
1940branches or with a lookup table.</p>
1941
1942<h5>Example:</h5>
1943
1944<pre>
1945 <i>; Emulate a conditional br instruction</i>
1946 %Val = <a href="#i_zext">zext</a> i1 %value to i32
1947 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
1948
1949 <i>; Emulate an unconditional br instruction</i>
1950 switch i32 0, label %dest [ ]
1951
1952 <i>; Implement a jump table:</i>
1953 switch i32 %val, label %otherwise [ i32 0, label %onzero
1954 i32 1, label %onone
1955 i32 2, label %ontwo ]
1956</pre>
1957</div>
1958
1959<!-- _______________________________________________________________________ -->
1960<div class="doc_subsubsection">
1961 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1962</div>
1963
1964<div class="doc_text">
1965
1966<h5>Syntax:</h5>
1967
1968<pre>
Nick Lewyckya1c11a12008-03-16 07:18:12 +0000[diff] [blame]1969 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] &lt;ptr to function ty&gt; &lt;function ptr val&gt;(&lt;function args&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1970 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
1971</pre>
1972
1973<h5>Overview:</h5>
1974
1975<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1976function, with the possibility of control flow transfer to either the
1977'<tt>normal</tt>' label or the
1978'<tt>exception</tt>' label. If the callee function returns with the
1979"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1980"normal" label. If the callee (or any indirect callees) returns with the "<a
1981href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]1982continued at the dynamically nearest "exception" label. If the callee function
Devang Patelec8a5b02008-03-11 05:51:59 +0000[diff] [blame]1983returns multiple values then individual return values are only accessible through
1984a '<tt><a href="#i_getresult">getresult</a></tt>' instruction.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1985
1986<h5>Arguments:</h5>
1987
1988<p>This instruction requires several arguments:</p>
1989
1990<ol>
1991 <li>
1992 The optional "cconv" marker indicates which <a href="#callingconv">calling
1993 convention</a> the call should use. If none is specified, the call defaults
1994 to using C calling conventions.
1995 </li>
1996 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1997 function value being invoked. In most cases, this is a direct function
1998 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1999 an arbitrary pointer to function value.
2000 </li>
2001
2002 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
2003 function to be invoked. </li>
2004
2005 <li>'<tt>function args</tt>': argument list whose types match the function
2006 signature argument types. If the function signature indicates the function
2007 accepts a variable number of arguments, the extra arguments can be
2008 specified. </li>
2009
2010 <li>'<tt>normal label</tt>': the label reached when the called function
2011 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
2012
2013 <li>'<tt>exception label</tt>': the label reached when a callee returns with
2014 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
2015
2016</ol>
2017
2018<h5>Semantics:</h5>
2019
2020<p>This instruction is designed to operate as a standard '<tt><a
2021href="#i_call">call</a></tt>' instruction in most regards. The primary
2022difference is that it establishes an association with a label, which is used by
2023the runtime library to unwind the stack.</p>
2024
2025<p>This instruction is used in languages with destructors to ensure that proper
2026cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
2027exception. Additionally, this is important for implementation of
2028'<tt>catch</tt>' clauses in high-level languages that support them.</p>
2029
2030<h5>Example:</h5>
2031<pre>
Nick Lewyckya1c11a12008-03-16 07:18:12 +0000[diff] [blame]2032 %retval = invoke i32 @Test(i32 15) to label %Continue
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2033 unwind label %TestCleanup <i>; {i32}:retval set</i>
Nick Lewyckya1c11a12008-03-16 07:18:12 +0000[diff] [blame]2034 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2035 unwind label %TestCleanup <i>; {i32}:retval set</i>
2036</pre>
2037</div>
2038
2039
2040<!-- _______________________________________________________________________ -->
2041
2042<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
2043Instruction</a> </div>
2044
2045<div class="doc_text">
2046
2047<h5>Syntax:</h5>
2048<pre>
2049 unwind
2050</pre>
2051
2052<h5>Overview:</h5>
2053
2054<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
2055at the first callee in the dynamic call stack which used an <a
2056href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
2057primarily used to implement exception handling.</p>
2058
2059<h5>Semantics:</h5>
2060
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]2061<p>The '<tt>unwind</tt>' instruction causes execution of the current function to
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2062immediately halt. The dynamic call stack is then searched for the first <a
2063href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
2064execution continues at the "exceptional" destination block specified by the
2065<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
2066dynamic call chain, undefined behavior results.</p>
2067</div>
2068
2069<!-- _______________________________________________________________________ -->
2070
2071<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
2072Instruction</a> </div>
2073
2074<div class="doc_text">
2075
2076<h5>Syntax:</h5>
2077<pre>
2078 unreachable
2079</pre>
2080
2081<h5>Overview:</h5>
2082
2083<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
2084instruction is used to inform the optimizer that a particular portion of the
2085code is not reachable. This can be used to indicate that the code after a
2086no-return function cannot be reached, and other facts.</p>
2087
2088<h5>Semantics:</h5>
2089
2090<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
2091</div>
2092
2093
2094
2095<!-- ======================================================================= -->
2096<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
2097<div class="doc_text">
2098<p>Binary operators are used to do most of the computation in a
Chris Lattnerab596d92008-04-01 18:47:32 +0000[diff] [blame]2099program. They require two operands of the same type, execute an operation on them, and
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2100produce a single value. The operands might represent
2101multiple 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]2102The result value has the same type as its operands.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2103<p>There are several different binary operators:</p>
2104</div>
2105<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2106<div class="doc_subsubsection">
2107 <a name="i_add">'<tt>add</tt>' Instruction</a>
2108</div>
2109
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2110<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2111
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2112<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2113
2114<pre>
2115 &lt;result&gt; = add &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2116</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2117
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2118<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2119
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2120<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2121
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2122<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2123
2124<p>The two arguments to the '<tt>add</tt>' instruction must be <a
2125 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>, or
2126 <a href="#t_vector">vector</a> values. Both arguments must have identical
2127 types.</p>
2128
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2129<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2130
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2131<p>The value produced is the integer or floating point sum of the two
2132operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2133
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2134<p>If an integer sum has unsigned overflow, the result returned is the
2135mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2136the result.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2137
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2138<p>Because LLVM integers use a two's complement representation, this
2139instruction is appropriate for both signed and unsigned integers.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2140
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2141<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2142
2143<pre>
2144 &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2145</pre>
2146</div>
2147<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2148<div class="doc_subsubsection">
2149 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
2150</div>
2151
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2152<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2153
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2154<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2155
2156<pre>
2157 &lt;result&gt; = sub &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2158</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2159
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2160<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2161
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2162<p>The '<tt>sub</tt>' instruction returns the difference of its two
2163operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2164
2165<p>Note that the '<tt>sub</tt>' instruction is used to represent the
2166'<tt>neg</tt>' instruction present in most other intermediate
2167representations.</p>
2168
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2169<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2170
2171<p>The two arguments to the '<tt>sub</tt>' instruction must be <a
2172 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2173 or <a href="#t_vector">vector</a> values. Both arguments must have identical
2174 types.</p>
2175
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2176<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2177
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2178<p>The value produced is the integer or floating point difference of
2179the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2180
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2181<p>If an integer difference has unsigned overflow, the result returned is the
2182mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2183the result.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2184
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2185<p>Because LLVM integers use a two's complement representation, this
2186instruction is appropriate for both signed and unsigned integers.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2187
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2188<h5>Example:</h5>
2189<pre>
2190 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
2191 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
2192</pre>
2193</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2194
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2195<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2196<div class="doc_subsubsection">
2197 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
2198</div>
2199
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2200<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2201
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2202<h5>Syntax:</h5>
2203<pre> &lt;result&gt; = mul &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2204</pre>
2205<h5>Overview:</h5>
2206<p>The '<tt>mul</tt>' instruction returns the product of its two
2207operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2208
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2209<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2210
2211<p>The two arguments to the '<tt>mul</tt>' instruction must be <a
2212href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2213or <a href="#t_vector">vector</a> values. Both arguments must have identical
2214types.</p>
2215
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2216<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2217
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2218<p>The value produced is the integer or floating point product of the
2219two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2220
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2221<p>If the result of an integer multiplication has unsigned overflow,
2222the result returned is the mathematical result modulo
22232<sup>n</sup>, where n is the bit width of the result.</p>
2224<p>Because LLVM integers use a two's complement representation, and the
2225result is the same width as the operands, this instruction returns the
2226correct result for both signed and unsigned integers. If a full product
2227(e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands
2228should be sign-extended or zero-extended as appropriate to the
2229width of the full product.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2230<h5>Example:</h5>
2231<pre> &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
2232</pre>
2233</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2234
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2235<!-- _______________________________________________________________________ -->
2236<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2237</a></div>
2238<div class="doc_text">
2239<h5>Syntax:</h5>
2240<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2241</pre>
2242<h5>Overview:</h5>
2243<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
2244operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2245
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2246<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2247
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2248<p>The two arguments to the '<tt>udiv</tt>' instruction must be
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2249<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2250values. Both arguments must have identical types.</p>
2251
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2252<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2253
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2254<p>The value produced is the unsigned integer quotient of the two operands.</p>
2255<p>Note that unsigned integer division and signed integer division are distinct
2256operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
2257<p>Division by zero leads to undefined behavior.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2258<h5>Example:</h5>
2259<pre> &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
2260</pre>
2261</div>
2262<!-- _______________________________________________________________________ -->
2263<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2264</a> </div>
2265<div class="doc_text">
2266<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2267<pre>
2268 &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2269</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2270
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2271<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2272
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2273<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
2274operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2275
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2276<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2277
2278<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
2279<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2280values. Both arguments must have identical types.</p>
2281
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2282<h5>Semantics:</h5>
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]2283<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]2284<p>Note that signed integer division and unsigned integer division are distinct
2285operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
2286<p>Division by zero leads to undefined behavior. Overflow also leads to
2287undefined behavior; this is a rare case, but can occur, for example,
2288by doing a 32-bit division of -2147483648 by -1.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2289<h5>Example:</h5>
2290<pre> &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
2291</pre>
2292</div>
2293<!-- _______________________________________________________________________ -->
2294<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
2295Instruction</a> </div>
2296<div class="doc_text">
2297<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2298<pre>
2299 &lt;result&gt; = fdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2300</pre>
2301<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2302
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2303<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
2304operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2305
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2306<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2307
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2308<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2309<a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2310of floating point values. Both arguments must have identical types.</p>
2311
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2312<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2313
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2314<p>The value produced is the floating point quotient of the two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2315
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2316<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2317
2318<pre>
2319 &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]2320</pre>
2321</div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2322
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2323<!-- _______________________________________________________________________ -->
2324<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2325</div>
2326<div class="doc_text">
2327<h5>Syntax:</h5>
2328<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2329</pre>
2330<h5>Overview:</h5>
2331<p>The '<tt>urem</tt>' instruction returns the remainder from the
2332unsigned division of its two arguments.</p>
2333<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2334<p>The two arguments to the '<tt>urem</tt>' instruction must be
2335<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2336values. Both arguments must have identical types.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2337<h5>Semantics:</h5>
2338<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]2339This instruction always performs an unsigned division to get the remainder.</p>
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2340<p>Note that unsigned integer remainder and signed integer remainder are
2341distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
2342<p>Taking the remainder of a division by zero leads to undefined behavior.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2343<h5>Example:</h5>
2344<pre> &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2345</pre>
2346
2347</div>
2348<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2349<div class="doc_subsubsection">
2350 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
2351</div>
2352
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2353<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2354
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2355<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2356
2357<pre>
2358 &lt;result&gt; = srem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2359</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2360
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2361<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2362
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2363<p>The '<tt>srem</tt>' instruction returns the remainder from the
Dan Gohman3e3fd8c2007-11-05 23:35:22 +0000[diff] [blame]2364signed division of its two operands. This instruction can also take
2365<a href="#t_vector">vector</a> versions of the values in which case
2366the elements must be integers.</p>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]2367
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2368<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2369
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2370<p>The two arguments to the '<tt>srem</tt>' instruction must be
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2371<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2372values. Both arguments must have identical types.</p>
2373
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2374<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2375
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2376<p>This instruction returns the <i>remainder</i> of a division (where the result
2377has the same sign as the dividend, <tt>var1</tt>), not the <i>modulo</i>
2378operator (where the result has the same sign as the divisor, <tt>var2</tt>) of
2379a value. For more information about the difference, see <a
2380 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
2381Math Forum</a>. For a table of how this is implemented in various languages,
2382please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
2383Wikipedia: modulo operation</a>.</p>
Chris Lattner9aba1e22008-01-28 00:36:27 +0000[diff] [blame]2384<p>Note that signed integer remainder and unsigned integer remainder are
2385distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
2386<p>Taking the remainder of a division by zero leads to undefined behavior.
2387Overflow also leads to undefined behavior; this is a rare case, but can occur,
2388for example, by taking the remainder of a 32-bit division of -2147483648 by -1.
2389(The remainder doesn't actually overflow, but this rule lets srem be
2390implemented using instructions that return both the result of the division
2391and the remainder.)</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2392<h5>Example:</h5>
2393<pre> &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2394</pre>
2395
2396</div>
2397<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2398<div class="doc_subsubsection">
2399 <a name="i_frem">'<tt>frem</tt>' Instruction</a> </div>
2400
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2401<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2402
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2403<h5>Syntax:</h5>
2404<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2405</pre>
2406<h5>Overview:</h5>
2407<p>The '<tt>frem</tt>' instruction returns the remainder from the
2408division of its two operands.</p>
2409<h5>Arguments:</h5>
2410<p>The two arguments to the '<tt>frem</tt>' instruction must be
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2411<a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2412of floating point values. Both arguments must have identical types.</p>
2413
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2414<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2415
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]2416<p>This instruction returns the <i>remainder</i> of a division.
2417The remainder has the same sign as the dividend.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2418
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2419<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2420
2421<pre>
2422 &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]2423</pre>
2424</div>
2425
2426<!-- ======================================================================= -->
2427<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2428Operations</a> </div>
2429<div class="doc_text">
2430<p>Bitwise binary operators are used to do various forms of
2431bit-twiddling in a program. They are generally very efficient
2432instructions and can commonly be strength reduced from other
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]2433instructions. They require two operands of the same type, execute an operation on them,
2434and 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]2435</div>
2436
2437<!-- _______________________________________________________________________ -->
2438<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2439Instruction</a> </div>
2440<div class="doc_text">
2441<h5>Syntax:</h5>
2442<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2443</pre>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2444
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2445<h5>Overview:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2446
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2447<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2448the left a specified number of bits.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2449
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2450<h5>Arguments:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2451
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2452<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]2453 href="#t_integer">integer</a> type. '<tt>var2</tt>' is treated as an
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2454unsigned value. This instruction does not support
2455<a href="#t_vector">vector</a> operands.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2456
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2457<h5>Semantics:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2458
Chris Lattnerdc15b1d2008-04-01 18:45:27 +0000[diff] [blame]2459<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup> mod 2<sup>n</sup>,
2460where n is the width of the result. If <tt>var2</tt> is (statically or dynamically) negative or
2461equal to or larger than the number of bits in <tt>var1</tt>, the result is undefined.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2462
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2463<h5>Example:</h5><pre>
2464 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2465 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2466 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2467 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2468</pre>
2469</div>
2470<!-- _______________________________________________________________________ -->
2471<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2472Instruction</a> </div>
2473<div class="doc_text">
2474<h5>Syntax:</h5>
2475<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2476</pre>
2477
2478<h5>Overview:</h5>
2479<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2480operand shifted to the right a specified number of bits with zero fill.</p>
2481
2482<h5>Arguments:</h5>
2483<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]2484<a href="#t_integer">integer</a> type. '<tt>var2</tt>' is treated as an
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2485unsigned value. This instruction does not support
2486<a href="#t_vector">vector</a> operands.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2487
2488<h5>Semantics:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2489
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2490<p>This instruction always performs a logical shift right operation. The most
2491significant bits of the result will be filled with zero bits after the
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2492shift. If <tt>var2</tt> is (statically or dynamically) equal to or larger than
2493the number of bits in <tt>var1</tt>, the result is undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2494
2495<h5>Example:</h5>
2496<pre>
2497 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2498 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2499 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2500 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2501 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2502</pre>
2503</div>
2504
2505<!-- _______________________________________________________________________ -->
2506<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2507Instruction</a> </div>
2508<div class="doc_text">
2509
2510<h5>Syntax:</h5>
2511<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2512</pre>
2513
2514<h5>Overview:</h5>
2515<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2516operand shifted to the right a specified number of bits with sign extension.</p>
2517
2518<h5>Arguments:</h5>
2519<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]2520<a href="#t_integer">integer</a> type. '<tt>var2</tt>' is treated as an
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2521unsigned value. This instruction does not support
2522<a href="#t_vector">vector</a> operands.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2523
2524<h5>Semantics:</h5>
2525<p>This instruction always performs an arithmetic shift right operation,
2526The most significant bits of the result will be filled with the sign bit
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2527of <tt>var1</tt>. If <tt>var2</tt> is (statically or dynamically) equal to or
2528larger than the number of bits in <tt>var1</tt>, the result is undefined.
2529</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2530
2531<h5>Example:</h5>
2532<pre>
2533 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2534 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2535 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2536 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2537 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2538</pre>
2539</div>
2540
2541<!-- _______________________________________________________________________ -->
2542<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2543Instruction</a> </div>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2544
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2545<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2546
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2547<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2548
2549<pre>
2550 &lt;result&gt; = and &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2551</pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2552
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2553<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2554
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2555<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2556its two operands.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2557
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2558<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2559
2560<p>The two arguments to the '<tt>and</tt>' instruction must be
2561<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2562values. Both arguments must have identical types.</p>
2563
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2564<h5>Semantics:</h5>
2565<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
2566<p> </p>
2567<div style="align: center">
2568<table border="1" cellspacing="0" cellpadding="4">
2569 <tbody>
2570 <tr>
2571 <td>In0</td>
2572 <td>In1</td>
2573 <td>Out</td>
2574 </tr>
2575 <tr>
2576 <td>0</td>
2577 <td>0</td>
2578 <td>0</td>
2579 </tr>
2580 <tr>
2581 <td>0</td>
2582 <td>1</td>
2583 <td>0</td>
2584 </tr>
2585 <tr>
2586 <td>1</td>
2587 <td>0</td>
2588 <td>0</td>
2589 </tr>
2590 <tr>
2591 <td>1</td>
2592 <td>1</td>
2593 <td>1</td>
2594 </tr>
2595 </tbody>
2596</table>
2597</div>
2598<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2599<pre>
2600 &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]2601 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2602 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
2603</pre>
2604</div>
2605<!-- _______________________________________________________________________ -->
2606<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
2607<div class="doc_text">
2608<h5>Syntax:</h5>
2609<pre> &lt;result&gt; = or &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2610</pre>
2611<h5>Overview:</h5>
2612<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2613or of its two operands.</p>
2614<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2615
2616<p>The two arguments to the '<tt>or</tt>' instruction must be
2617<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2618values. Both arguments must have identical types.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2619<h5>Semantics:</h5>
2620<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
2621<p> </p>
2622<div style="align: center">
2623<table border="1" cellspacing="0" cellpadding="4">
2624 <tbody>
2625 <tr>
2626 <td>In0</td>
2627 <td>In1</td>
2628 <td>Out</td>
2629 </tr>
2630 <tr>
2631 <td>0</td>
2632 <td>0</td>
2633 <td>0</td>
2634 </tr>
2635 <tr>
2636 <td>0</td>
2637 <td>1</td>
2638 <td>1</td>
2639 </tr>
2640 <tr>
2641 <td>1</td>
2642 <td>0</td>
2643 <td>1</td>
2644 </tr>
2645 <tr>
2646 <td>1</td>
2647 <td>1</td>
2648 <td>1</td>
2649 </tr>
2650 </tbody>
2651</table>
2652</div>
2653<h5>Example:</h5>
2654<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2655 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2656 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
2657</pre>
2658</div>
2659<!-- _______________________________________________________________________ -->
2660<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2661Instruction</a> </div>
2662<div class="doc_text">
2663<h5>Syntax:</h5>
2664<pre> &lt;result&gt; = xor &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2665</pre>
2666<h5>Overview:</h5>
2667<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2668or of its two operands. The <tt>xor</tt> is used to implement the
2669"one's complement" operation, which is the "~" operator in C.</p>
2670<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2671<p>The two arguments to the '<tt>xor</tt>' instruction must be
2672<a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2673values. Both arguments must have identical types.</p>
2674
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2675<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]2676
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2677<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
2678<p> </p>
2679<div style="align: center">
2680<table border="1" cellspacing="0" cellpadding="4">
2681 <tbody>
2682 <tr>
2683 <td>In0</td>
2684 <td>In1</td>
2685 <td>Out</td>
2686 </tr>
2687 <tr>
2688 <td>0</td>
2689 <td>0</td>
2690 <td>0</td>
2691 </tr>
2692 <tr>
2693 <td>0</td>
2694 <td>1</td>
2695 <td>1</td>
2696 </tr>
2697 <tr>
2698 <td>1</td>
2699 <td>0</td>
2700 <td>1</td>
2701 </tr>
2702 <tr>
2703 <td>1</td>
2704 <td>1</td>
2705 <td>0</td>
2706 </tr>
2707 </tbody>
2708</table>
2709</div>
2710<p> </p>
2711<h5>Example:</h5>
2712<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2713 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2714 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2715 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
2716</pre>
2717</div>
2718
2719<!-- ======================================================================= -->
2720<div class="doc_subsection">
2721 <a name="vectorops">Vector Operations</a>
2722</div>
2723
2724<div class="doc_text">
2725
2726<p>LLVM supports several instructions to represent vector operations in a
2727target-independent manner. These instructions cover the element-access and
2728vector-specific operations needed to process vectors effectively. While LLVM
2729does directly support these vector operations, many sophisticated algorithms
2730will want to use target-specific intrinsics to take full advantage of a specific
2731target.</p>
2732
2733</div>
2734
2735<!-- _______________________________________________________________________ -->
2736<div class="doc_subsubsection">
2737 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2738</div>
2739
2740<div class="doc_text">
2741
2742<h5>Syntax:</h5>
2743
2744<pre>
2745 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
2746</pre>
2747
2748<h5>Overview:</h5>
2749
2750<p>
2751The '<tt>extractelement</tt>' instruction extracts a single scalar
2752element from a vector at a specified index.
2753</p>
2754
2755
2756<h5>Arguments:</h5>
2757
2758<p>
2759The first operand of an '<tt>extractelement</tt>' instruction is a
2760value of <a href="#t_vector">vector</a> type. The second operand is
2761an index indicating the position from which to extract the element.
2762The index may be a variable.</p>
2763
2764<h5>Semantics:</h5>
2765
2766<p>
2767The result is a scalar of the same type as the element type of
2768<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2769<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2770results are undefined.
2771</p>
2772
2773<h5>Example:</h5>
2774
2775<pre>
2776 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
2777</pre>
2778</div>
2779
2780
2781<!-- _______________________________________________________________________ -->
2782<div class="doc_subsubsection">
2783 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2784</div>
2785
2786<div class="doc_text">
2787
2788<h5>Syntax:</h5>
2789
2790<pre>
Dan Gohmanbcc3c502008-05-12 23:38:42 +0000[diff] [blame]2791 &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]2792</pre>
2793
2794<h5>Overview:</h5>
2795
2796<p>
2797The '<tt>insertelement</tt>' instruction inserts a scalar
2798element into a vector at a specified index.
2799</p>
2800
2801
2802<h5>Arguments:</h5>
2803
2804<p>
2805The first operand of an '<tt>insertelement</tt>' instruction is a
2806value of <a href="#t_vector">vector</a> type. The second operand is a
2807scalar value whose type must equal the element type of the first
2808operand. The third operand is an index indicating the position at
2809which to insert the value. The index may be a variable.</p>
2810
2811<h5>Semantics:</h5>
2812
2813<p>
2814The result is a vector of the same type as <tt>val</tt>. Its
2815element values are those of <tt>val</tt> except at position
2816<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2817exceeds the length of <tt>val</tt>, the results are undefined.
2818</p>
2819
2820<h5>Example:</h5>
2821
2822<pre>
2823 %result = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
2824</pre>
2825</div>
2826
2827<!-- _______________________________________________________________________ -->
2828<div class="doc_subsubsection">
2829 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2830</div>
2831
2832<div class="doc_text">
2833
2834<h5>Syntax:</h5>
2835
2836<pre>
2837 &lt;result&gt; = shufflevector &lt;n x &lt;ty&gt;&gt; &lt;v1&gt;, &lt;n x &lt;ty&gt;&gt; &lt;v2&gt;, &lt;n x i32&gt; &lt;mask&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
2838</pre>
2839
2840<h5>Overview:</h5>
2841
2842<p>
2843The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2844from two input vectors, returning a vector of the same type.
2845</p>
2846
2847<h5>Arguments:</h5>
2848
2849<p>
2850The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2851with types that match each other and types that match the result of the
2852instruction. The third argument is a shuffle mask, which has the same number
2853of elements as the other vector type, but whose element type is always 'i32'.
2854</p>
2855
2856<p>
2857The shuffle mask operand is required to be a constant vector with either
2858constant integer or undef values.
2859</p>
2860
2861<h5>Semantics:</h5>
2862
2863<p>
2864The elements of the two input vectors are numbered from left to right across
2865both of the vectors. The shuffle mask operand specifies, for each element of
2866the result vector, which element of the two input registers the result element
2867gets. The element selector may be undef (meaning "don't care") and the second
2868operand may be undef if performing a shuffle from only one vector.
2869</p>
2870
2871<h5>Example:</h5>
2872
2873<pre>
2874 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
2875 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
2876 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
2877 &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.
2878</pre>
2879</div>
2880
2881
2882<!-- ======================================================================= -->
2883<div class="doc_subsection">
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]2884 <a name="aggregateops">Aggregate Operations</a>
2885</div>
2886
2887<div class="doc_text">
2888
2889<p>LLVM supports several instructions for working with aggregate values.
2890</p>
2891
2892</div>
2893
2894<!-- _______________________________________________________________________ -->
2895<div class="doc_subsubsection">
2896 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
2897</div>
2898
2899<div class="doc_text">
2900
2901<h5>Syntax:</h5>
2902
2903<pre>
2904 &lt;result&gt; = extractvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;idx&gt;{, &lt;idx&gt;}*
2905</pre>
2906
2907<h5>Overview:</h5>
2908
2909<p>
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]2910The '<tt>extractvalue</tt>' instruction extracts the value of a struct field
2911or array element from an aggregate value.
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]2912</p>
2913
2914
2915<h5>Arguments:</h5>
2916
2917<p>
2918The first operand of an '<tt>extractvalue</tt>' instruction is a
2919value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a>
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]2920type. The operands are constant indices to specify which value to extract
2921in the same manner as indices in a
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]2922'<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
2923</p>
2924
2925<h5>Semantics:</h5>
2926
2927<p>
2928The result is the value at the position in the aggregate specified by
2929the index operands.
2930</p>
2931
2932<h5>Example:</h5>
2933
2934<pre>
2935 %result = extractvalue {i32, float} %agg, i32 0 <i>; yields i32</i>
2936</pre>
2937</div>
2938
2939
2940<!-- _______________________________________________________________________ -->
2941<div class="doc_subsubsection">
2942 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
2943</div>
2944
2945<div class="doc_text">
2946
2947<h5>Syntax:</h5>
2948
2949<pre>
2950 &lt;result&gt; = insertvalue &lt;aggregate type&gt; &lt;val&gt;, &lt;ty&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;n x &lt;ty&gt;&gt;</i>
2951</pre>
2952
2953<h5>Overview:</h5>
2954
2955<p>
2956The '<tt>insertvalue</tt>' instruction inserts a value
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]2957into a struct field or array element in an aggregate.
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]2958</p>
2959
2960
2961<h5>Arguments:</h5>
2962
2963<p>
2964The first operand of an '<tt>insertvalue</tt>' instruction is a
2965value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a> type.
2966The second operand is a first-class value to insert.
Dan Gohman4f29e422008-05-23 21:53:15 +0000[diff] [blame]2967The following operands are constant indices
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]2968indicating the position at which to insert the value in the same manner as
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]2969indices in a
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]2970'<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
2971The value to insert must have the same type as the value identified
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]2972by the indices.
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]2973
2974<h5>Semantics:</h5>
2975
2976<p>
2977The result is an aggregate of the same type as <tt>val</tt>. Its
2978value is that of <tt>val</tt> except that the value at the position
Dan Gohman6c6dea02008-05-13 18:16:06 +0000[diff] [blame]2979specified by the indices is that of <tt>elt</tt>.
Dan Gohman74d6faf2008-05-12 23:51:09 +0000[diff] [blame]2980</p>
2981
2982<h5>Example:</h5>
2983
2984<pre>
2985 %result = insertvalue {i32, float} %agg, i32 1, i32 0 <i>; yields {i32, float}</i>
2986</pre>
2987</div>
2988
2989
2990<!-- ======================================================================= -->
2991<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2992 <a name="memoryops">Memory Access and Addressing Operations</a>
2993</div>
2994
2995<div class="doc_text">
2996
2997<p>A key design point of an SSA-based representation is how it
2998represents memory. In LLVM, no memory locations are in SSA form, which
2999makes things very simple. This section describes how to read, write,
3000allocate, and free memory in LLVM.</p>
3001
3002</div>
3003
3004<!-- _______________________________________________________________________ -->
3005<div class="doc_subsubsection">
3006 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
3007</div>
3008
3009<div class="doc_text">
3010
3011<h5>Syntax:</h5>
3012
3013<pre>
3014 &lt;result&gt; = malloc &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
3015</pre>
3016
3017<h5>Overview:</h5>
3018
3019<p>The '<tt>malloc</tt>' instruction allocates memory from the system
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]3020heap and returns a pointer to it. The object is always allocated in the generic
3021address space (address space zero).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3022
3023<h5>Arguments:</h5>
3024
3025<p>The '<tt>malloc</tt>' instruction allocates
3026<tt>sizeof(&lt;type&gt;)*NumElements</tt>
3027bytes of memory from the operating system and returns a pointer of the
3028appropriate type to the program. If "NumElements" is specified, it is the
Gabor Greif5082cf42008-02-09 22:24:34 +0000[diff] [blame]3029number of elements allocated, otherwise "NumElements" is defaulted to be one.
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3030If a constant alignment is specified, the value result of the allocation is guaranteed to
Gabor Greif5082cf42008-02-09 22:24:34 +0000[diff] [blame]3031be aligned to at least that boundary. If not specified, or if zero, the target can
3032choose to align the allocation on any convenient boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3033
3034<p>'<tt>type</tt>' must be a sized type.</p>
3035
3036<h5>Semantics:</h5>
3037
3038<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]3039a pointer is returned. The result of a zero byte allocattion is undefined. The
3040result is null if there is insufficient memory available.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3041
3042<h5>Example:</h5>
3043
3044<pre>
3045 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
3046
3047 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
3048 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
3049 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
3050 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
3051 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
3052</pre>
3053</div>
3054
3055<!-- _______________________________________________________________________ -->
3056<div class="doc_subsubsection">
3057 <a name="i_free">'<tt>free</tt>' Instruction</a>
3058</div>
3059
3060<div class="doc_text">
3061
3062<h5>Syntax:</h5>
3063
3064<pre>
3065 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
3066</pre>
3067
3068<h5>Overview:</h5>
3069
3070<p>The '<tt>free</tt>' instruction returns memory back to the unused
3071memory heap to be reallocated in the future.</p>
3072
3073<h5>Arguments:</h5>
3074
3075<p>'<tt>value</tt>' shall be a pointer value that points to a value
3076that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
3077instruction.</p>
3078
3079<h5>Semantics:</h5>
3080
3081<p>Access to the memory pointed to by the pointer is no longer defined
Chris Lattner329d6532008-04-19 22:41:32 +0000[diff] [blame]3082after this instruction executes. If the pointer is null, the operation
3083is a noop.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3084
3085<h5>Example:</h5>
3086
3087<pre>
3088 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
3089 free [4 x i8]* %array
3090</pre>
3091</div>
3092
3093<!-- _______________________________________________________________________ -->
3094<div class="doc_subsubsection">
3095 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
3096</div>
3097
3098<div class="doc_text">
3099
3100<h5>Syntax:</h5>
3101
3102<pre>
3103 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
3104</pre>
3105
3106<h5>Overview:</h5>
3107
3108<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
3109currently executing function, to be automatically released when this function
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]3110returns to its caller. The object is always allocated in the generic address
3111space (address space zero).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3112
3113<h5>Arguments:</h5>
3114
3115<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
3116bytes of memory on the runtime stack, returning a pointer of the
Gabor Greif5082cf42008-02-09 22:24:34 +0000[diff] [blame]3117appropriate type to the program. If "NumElements" is specified, it is the
3118number of elements allocated, otherwise "NumElements" is defaulted to be one.
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3119If a constant alignment is specified, the value result of the allocation is guaranteed
Gabor Greif5082cf42008-02-09 22:24:34 +0000[diff] [blame]3120to be aligned to at least that boundary. If not specified, or if zero, the target
3121can choose to align the allocation on any convenient boundary.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3122
3123<p>'<tt>type</tt>' may be any sized type.</p>
3124
3125<h5>Semantics:</h5>
3126
Chris Lattner8b094fc2008-04-19 21:01:16 +0000[diff] [blame]3127<p>Memory is allocated; a pointer is returned. The operation is undefiend if
3128there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3129memory is automatically released when the function returns. The '<tt>alloca</tt>'
3130instruction is commonly used to represent automatic variables that must
3131have an address available. When the function returns (either with the <tt><a
3132 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]3133instructions), the memory is reclaimed. Allocating zero bytes
3134is legal, but the result is undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3135
3136<h5>Example:</h5>
3137
3138<pre>
3139 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
3140 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
3141 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
3142 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
3143</pre>
3144</div>
3145
3146<!-- _______________________________________________________________________ -->
3147<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
3148Instruction</a> </div>
3149<div class="doc_text">
3150<h5>Syntax:</h5>
3151<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>
3152<h5>Overview:</h5>
3153<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
3154<h5>Arguments:</h5>
3155<p>The argument to the '<tt>load</tt>' instruction specifies the memory
3156address from which to load. The pointer must point to a <a
3157 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
3158marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
3159the number or order of execution of this <tt>load</tt> with other
3160volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
3161instructions. </p>
Chris Lattner21003c82008-01-06 21:04:43 +0000[diff] [blame]3162<p>
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3163The optional constant "align" argument specifies the alignment of the operation
Chris Lattner21003c82008-01-06 21:04:43 +0000[diff] [blame]3164(that is, the alignment of the memory address). A value of 0 or an
3165omitted "align" argument means that the operation has the preferential
3166alignment for the target. It is the responsibility of the code emitter
3167to ensure that the alignment information is correct. Overestimating
3168the alignment results in an undefined behavior. Underestimating the
3169alignment may produce less efficient code. An alignment of 1 is always
3170safe.
3171</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3172<h5>Semantics:</h5>
3173<p>The location of memory pointed to is loaded.</p>
3174<h5>Examples:</h5>
3175<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
3176 <a
3177 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
3178 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
3179</pre>
3180</div>
3181<!-- _______________________________________________________________________ -->
3182<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
3183Instruction</a> </div>
3184<div class="doc_text">
3185<h5>Syntax:</h5>
3186<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
3187 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
3188</pre>
3189<h5>Overview:</h5>
3190<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
3191<h5>Arguments:</h5>
3192<p>There are two arguments to the '<tt>store</tt>' instruction: a value
3193to 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]3194operand must be a pointer to the <a href="#t_firstclass">first class</a> type
3195of the '<tt>&lt;value&gt;</tt>'
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3196operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
3197optimizer is not allowed to modify the number or order of execution of
3198this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
3199 href="#i_store">store</a></tt> instructions.</p>
Chris Lattner21003c82008-01-06 21:04:43 +0000[diff] [blame]3200<p>
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3201The optional constant "align" argument specifies the alignment of the operation
Chris Lattner21003c82008-01-06 21:04:43 +0000[diff] [blame]3202(that is, the alignment of the memory address). A value of 0 or an
3203omitted "align" argument means that the operation has the preferential
3204alignment for the target. It is the responsibility of the code emitter
3205to ensure that the alignment information is correct. Overestimating
3206the alignment results in an undefined behavior. Underestimating the
3207alignment may produce less efficient code. An alignment of 1 is always
3208safe.
3209</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3210<h5>Semantics:</h5>
3211<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
3212at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
3213<h5>Example:</h5>
3214<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling63ffa142007-10-22 05:10:05 +0000[diff] [blame]3215 store i32 3, i32* %ptr <i>; yields {void}</i>
3216 %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]3217</pre>
3218</div>
3219
3220<!-- _______________________________________________________________________ -->
3221<div class="doc_subsubsection">
3222 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
3223</div>
3224
3225<div class="doc_text">
3226<h5>Syntax:</h5>
3227<pre>
3228 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
3229</pre>
3230
3231<h5>Overview:</h5>
3232
3233<p>
3234The '<tt>getelementptr</tt>' instruction is used to get the address of a
3235subelement of an aggregate data structure.</p>
3236
3237<h5>Arguments:</h5>
3238
3239<p>This instruction takes a list of integer operands that indicate what
3240elements of the aggregate object to index to. The actual types of the arguments
3241provided depend on the type of the first pointer argument. The
3242'<tt>getelementptr</tt>' instruction is used to index down through the type
3243levels of a structure or to a specific index in an array. When indexing into a
3244structure, only <tt>i32</tt> integer constants are allowed. When indexing
Chris Lattnera7d94ba2008-04-24 05:59:56 +0000[diff] [blame]3245into an array or pointer, only integers of 32 or 64 bits are allowed; 32-bit
3246values will be sign extended to 64-bits if required.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3247
3248<p>For example, let's consider a C code fragment and how it gets
3249compiled to LLVM:</p>
3250
3251<div class="doc_code">
3252<pre>
3253struct RT {
3254 char A;
3255 int B[10][20];
3256 char C;
3257};
3258struct ST {
3259 int X;
3260 double Y;
3261 struct RT Z;
3262};
3263
3264int *foo(struct ST *s) {
3265 return &amp;s[1].Z.B[5][13];
3266}
3267</pre>
3268</div>
3269
3270<p>The LLVM code generated by the GCC frontend is:</p>
3271
3272<div class="doc_code">
3273<pre>
3274%RT = type { i8 , [10 x [20 x i32]], i8 }
3275%ST = type { i32, double, %RT }
3276
3277define i32* %foo(%ST* %s) {
3278entry:
3279 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
3280 ret i32* %reg
3281}
3282</pre>
3283</div>
3284
3285<h5>Semantics:</h5>
3286
3287<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
3288on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
3289and <a href="#t_array">array</a> types can use a 32-bit or 64-bit
3290<a href="#t_integer">integer</a> type but the value will always be sign extended
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3291to 64-bits. <a href="#t_struct">Structure</a> and <a href="#t_pstruct">packed
3292structure</a> types require <tt>i32</tt> <b>constants</b>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3293
3294<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
3295type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
3296}</tt>' type, a structure. The second index indexes into the third element of
3297the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
3298i8 }</tt>' type, another structure. The third index indexes into the second
3299element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
3300array. The two dimensions of the array are subscripted into, yielding an
3301'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
3302to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
3303
3304<p>Note that it is perfectly legal to index partially through a
3305structure, returning a pointer to an inner element. Because of this,
3306the LLVM code for the given testcase is equivalent to:</p>
3307
3308<pre>
3309 define i32* %foo(%ST* %s) {
3310 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
3311 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
3312 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
3313 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
3314 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
3315 ret i32* %t5
3316 }
3317</pre>
3318
3319<p>Note that it is undefined to access an array out of bounds: array and
3320pointer indexes must always be within the defined bounds of the array type.
Chris Lattnera7d94ba2008-04-24 05:59:56 +0000[diff] [blame]3321The one exception for this rule is zero length arrays. These arrays are
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3322defined to be accessible as variable length arrays, which requires access
3323beyond the zero'th element.</p>
3324
3325<p>The getelementptr instruction is often confusing. For some more insight
3326into how it works, see <a href="GetElementPtr.html">the getelementptr
3327FAQ</a>.</p>
3328
3329<h5>Example:</h5>
3330
3331<pre>
3332 <i>; yields [12 x i8]*:aptr</i>
3333 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
3334</pre>
3335</div>
3336
3337<!-- ======================================================================= -->
3338<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
3339</div>
3340<div class="doc_text">
3341<p>The instructions in this category are the conversion instructions (casting)
3342which all take a single operand and a type. They perform various bit conversions
3343on the operand.</p>
3344</div>
3345
3346<!-- _______________________________________________________________________ -->
3347<div class="doc_subsubsection">
3348 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
3349</div>
3350<div class="doc_text">
3351
3352<h5>Syntax:</h5>
3353<pre>
3354 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3355</pre>
3356
3357<h5>Overview:</h5>
3358<p>
3359The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
3360</p>
3361
3362<h5>Arguments:</h5>
3363<p>
3364The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
3365be an <a href="#t_integer">integer</a> type, and a type that specifies the size
3366and type of the result, which must be an <a href="#t_integer">integer</a>
3367type. The bit size of <tt>value</tt> must be larger than the bit size of
3368<tt>ty2</tt>. Equal sized types are not allowed.</p>
3369
3370<h5>Semantics:</h5>
3371<p>
3372The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
3373and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
3374larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
3375It will always truncate bits.</p>
3376
3377<h5>Example:</h5>
3378<pre>
3379 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
3380 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
3381 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
3382</pre>
3383</div>
3384
3385<!-- _______________________________________________________________________ -->
3386<div class="doc_subsubsection">
3387 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
3388</div>
3389<div class="doc_text">
3390
3391<h5>Syntax:</h5>
3392<pre>
3393 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3394</pre>
3395
3396<h5>Overview:</h5>
3397<p>The '<tt>zext</tt>' instruction zero extends its operand to type
3398<tt>ty2</tt>.</p>
3399
3400
3401<h5>Arguments:</h5>
3402<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
3403<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3404also be of <a href="#t_integer">integer</a> type. The bit size of the
3405<tt>value</tt> must be smaller than the bit size of the destination type,
3406<tt>ty2</tt>.</p>
3407
3408<h5>Semantics:</h5>
3409<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
3410bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
3411
3412<p>When zero extending from i1, the result will always be either 0 or 1.</p>
3413
3414<h5>Example:</h5>
3415<pre>
3416 %X = zext i32 257 to i64 <i>; yields i64:257</i>
3417 %Y = zext i1 true to i32 <i>; yields i32:1</i>
3418</pre>
3419</div>
3420
3421<!-- _______________________________________________________________________ -->
3422<div class="doc_subsubsection">
3423 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
3424</div>
3425<div class="doc_text">
3426
3427<h5>Syntax:</h5>
3428<pre>
3429 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3430</pre>
3431
3432<h5>Overview:</h5>
3433<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
3434
3435<h5>Arguments:</h5>
3436<p>
3437The '<tt>sext</tt>' instruction takes a value to cast, which must be of
3438<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3439also be of <a href="#t_integer">integer</a> type. The bit size of the
3440<tt>value</tt> must be smaller than the bit size of the destination type,
3441<tt>ty2</tt>.</p>
3442
3443<h5>Semantics:</h5>
3444<p>
3445The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
3446bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
3447the type <tt>ty2</tt>.</p>
3448
3449<p>When sign extending from i1, the extension always results in -1 or 0.</p>
3450
3451<h5>Example:</h5>
3452<pre>
3453 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
3454 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
3455</pre>
3456</div>
3457
3458<!-- _______________________________________________________________________ -->
3459<div class="doc_subsubsection">
3460 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3461</div>
3462
3463<div class="doc_text">
3464
3465<h5>Syntax:</h5>
3466
3467<pre>
3468 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3469</pre>
3470
3471<h5>Overview:</h5>
3472<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
3473<tt>ty2</tt>.</p>
3474
3475
3476<h5>Arguments:</h5>
3477<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
3478 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
3479cast it to. The size of <tt>value</tt> must be larger than the size of
3480<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3481<i>no-op cast</i>.</p>
3482
3483<h5>Semantics:</h5>
3484<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3485<a href="#t_floating">floating point</a> type to a smaller
3486<a href="#t_floating">floating point</a> type. If the value cannot fit within
3487the destination type, <tt>ty2</tt>, then the results are undefined.</p>
3488
3489<h5>Example:</h5>
3490<pre>
3491 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3492 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3493</pre>
3494</div>
3495
3496<!-- _______________________________________________________________________ -->
3497<div class="doc_subsubsection">
3498 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
3499</div>
3500<div class="doc_text">
3501
3502<h5>Syntax:</h5>
3503<pre>
3504 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3505</pre>
3506
3507<h5>Overview:</h5>
3508<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
3509floating point value.</p>
3510
3511<h5>Arguments:</h5>
3512<p>The '<tt>fpext</tt>' instruction takes a
3513<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
3514and a <a href="#t_floating">floating point</a> type to cast it to. The source
3515type must be smaller than the destination type.</p>
3516
3517<h5>Semantics:</h5>
3518<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
3519<a href="#t_floating">floating point</a> type to a larger
3520<a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
3521used to make a <i>no-op cast</i> because it always changes bits. Use
3522<tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
3523
3524<h5>Example:</h5>
3525<pre>
3526 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
3527 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
3528</pre>
3529</div>
3530
3531<!-- _______________________________________________________________________ -->
3532<div class="doc_subsubsection">
3533 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
3534</div>
3535<div class="doc_text">
3536
3537<h5>Syntax:</h5>
3538<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3539 &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]3540</pre>
3541
3542<h5>Overview:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3543<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3544unsigned integer equivalent of type <tt>ty2</tt>.
3545</p>
3546
3547<h5>Arguments:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3548<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]3549scalar or vector <a href="#t_floating">floating point</a> value, and a type
3550to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3551type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3552vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3553
3554<h5>Semantics:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3555<p> The '<tt>fptoui</tt>' instruction converts its
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3556<a href="#t_floating">floating point</a> operand into the nearest (rounding
3557towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3558the results are undefined.</p>
3559
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3560<h5>Example:</h5>
3561<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3562 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]3563 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3564 %X = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3565</pre>
3566</div>
3567
3568<!-- _______________________________________________________________________ -->
3569<div class="doc_subsubsection">
3570 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
3571</div>
3572<div class="doc_text">
3573
3574<h5>Syntax:</h5>
3575<pre>
3576 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3577</pre>
3578
3579<h5>Overview:</h5>
3580<p>The '<tt>fptosi</tt>' instruction converts
3581<a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
3582</p>
3583
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3584<h5>Arguments:</h5>
3585<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]3586scalar or vector <a href="#t_floating">floating point</a> value, and a type
3587to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3588type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3589vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3590
3591<h5>Semantics:</h5>
3592<p>The '<tt>fptosi</tt>' instruction converts its
3593<a href="#t_floating">floating point</a> operand into the nearest (rounding
3594towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3595the results are undefined.</p>
3596
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3597<h5>Example:</h5>
3598<pre>
3599 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]3600 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3601 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
3602</pre>
3603</div>
3604
3605<!-- _______________________________________________________________________ -->
3606<div class="doc_subsubsection">
3607 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
3608</div>
3609<div class="doc_text">
3610
3611<h5>Syntax:</h5>
3612<pre>
3613 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3614</pre>
3615
3616<h5>Overview:</h5>
3617<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
3618integer and converts that value to the <tt>ty2</tt> type.</p>
3619
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3620<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]3621<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a
3622scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3623to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3624type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3625floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3626
3627<h5>Semantics:</h5>
3628<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
3629integer quantity and converts it to the corresponding floating point value. If
3630the value cannot fit in the floating point value, the results are undefined.</p>
3631
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3632<h5>Example:</h5>
3633<pre>
3634 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
3635 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
3636</pre>
3637</div>
3638
3639<!-- _______________________________________________________________________ -->
3640<div class="doc_subsubsection">
3641 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
3642</div>
3643<div class="doc_text">
3644
3645<h5>Syntax:</h5>
3646<pre>
3647 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3648</pre>
3649
3650<h5>Overview:</h5>
3651<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
3652integer and converts that value to the <tt>ty2</tt> type.</p>
3653
3654<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]3655<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a
3656scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3657to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3658type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3659floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3660
3661<h5>Semantics:</h5>
3662<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
3663integer quantity and converts it to the corresponding floating point value. If
3664the value cannot fit in the floating point value, the results are undefined.</p>
3665
3666<h5>Example:</h5>
3667<pre>
3668 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
3669 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
3670</pre>
3671</div>
3672
3673<!-- _______________________________________________________________________ -->
3674<div class="doc_subsubsection">
3675 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3676</div>
3677<div class="doc_text">
3678
3679<h5>Syntax:</h5>
3680<pre>
3681 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3682</pre>
3683
3684<h5>Overview:</h5>
3685<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3686the integer type <tt>ty2</tt>.</p>
3687
3688<h5>Arguments:</h5>
3689<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
3690must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
3691<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3692
3693<h5>Semantics:</h5>
3694<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3695<tt>ty2</tt> by interpreting the pointer value as an integer and either
3696truncating or zero extending that value to the size of the integer type. If
3697<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3698<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
3699are the same size, then nothing is done (<i>no-op cast</i>) other than a type
3700change.</p>
3701
3702<h5>Example:</h5>
3703<pre>
3704 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
3705 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
3706</pre>
3707</div>
3708
3709<!-- _______________________________________________________________________ -->
3710<div class="doc_subsubsection">
3711 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3712</div>
3713<div class="doc_text">
3714
3715<h5>Syntax:</h5>
3716<pre>
3717 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3718</pre>
3719
3720<h5>Overview:</h5>
3721<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3722a pointer type, <tt>ty2</tt>.</p>
3723
3724<h5>Arguments:</h5>
3725<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
3726value to cast, and a type to cast it to, which must be a
3727<a href="#t_pointer">pointer</a> type.
3728
3729<h5>Semantics:</h5>
3730<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3731<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3732the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3733size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3734the size of a pointer then a zero extension is done. If they are the same size,
3735nothing is done (<i>no-op cast</i>).</p>
3736
3737<h5>Example:</h5>
3738<pre>
3739 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
3740 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
3741 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
3742</pre>
3743</div>
3744
3745<!-- _______________________________________________________________________ -->
3746<div class="doc_subsubsection">
3747 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
3748</div>
3749<div class="doc_text">
3750
3751<h5>Syntax:</h5>
3752<pre>
3753 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3754</pre>
3755
3756<h5>Overview:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3757
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3758<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
3759<tt>ty2</tt> without changing any bits.</p>
3760
3761<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3762
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3763<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
3764a first class value, and a type to cast it to, which must also be a <a
3765 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
3766and the destination type, <tt>ty2</tt>, must be identical. If the source
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3767type is a pointer, the destination type must also be a pointer. This
3768instruction supports bitwise conversion of vectors to integers and to vectors
3769of other types (as long as they have the same size).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3770
3771<h5>Semantics:</h5>
3772<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
3773<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3774this conversion. The conversion is done as if the <tt>value</tt> had been
3775stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3776converted to other pointer types with this instruction. To convert pointers to
3777other types, use the <a href="#i_inttoptr">inttoptr</a> or
3778<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
3779
3780<h5>Example:</h5>
3781<pre>
3782 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
3783 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3784 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
3785</pre>
3786</div>
3787
3788<!-- ======================================================================= -->
3789<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3790<div class="doc_text">
3791<p>The instructions in this category are the "miscellaneous"
3792instructions, which defy better classification.</p>
3793</div>
3794
3795<!-- _______________________________________________________________________ -->
3796<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3797</div>
3798<div class="doc_text">
3799<h5>Syntax:</h5>
3800<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
3801</pre>
3802<h5>Overview:</h5>
3803<p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
Chris Lattner10368b62008-04-02 00:38:26 +0000[diff] [blame]3804of its two integer or pointer operands.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3805<h5>Arguments:</h5>
3806<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
3807the condition code indicating the kind of comparison to perform. It is not
3808a value, just a keyword. The possible condition code are:
3809<ol>
3810 <li><tt>eq</tt>: equal</li>
3811 <li><tt>ne</tt>: not equal </li>
3812 <li><tt>ugt</tt>: unsigned greater than</li>
3813 <li><tt>uge</tt>: unsigned greater or equal</li>
3814 <li><tt>ult</tt>: unsigned less than</li>
3815 <li><tt>ule</tt>: unsigned less or equal</li>
3816 <li><tt>sgt</tt>: signed greater than</li>
3817 <li><tt>sge</tt>: signed greater or equal</li>
3818 <li><tt>slt</tt>: signed less than</li>
3819 <li><tt>sle</tt>: signed less or equal</li>
3820</ol>
3821<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
3822<a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
3823<h5>Semantics:</h5>
3824<p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3825the condition code given as <tt>cond</tt>. The comparison performed always
3826yields a <a href="#t_primitive">i1</a> result, as follows:
3827<ol>
3828 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3829 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3830 </li>
3831 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3832 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3833 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3834 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3835 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3836 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3837 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3838 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3839 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3840 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3841 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3842 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3843 <li><tt>sge</tt>: interprets the operands as signed values and yields
3844 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3845 <li><tt>slt</tt>: interprets the operands as signed values and yields
3846 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3847 <li><tt>sle</tt>: interprets the operands as signed values and yields
3848 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3849</ol>
3850<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
3851values are compared as if they were integers.</p>
3852
3853<h5>Example:</h5>
3854<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
3855 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
3856 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
3857 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3858 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3859 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
3860</pre>
3861</div>
3862
3863<!-- _______________________________________________________________________ -->
3864<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3865</div>
3866<div class="doc_text">
3867<h5>Syntax:</h5>
3868<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
3869</pre>
3870<h5>Overview:</h5>
3871<p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3872of its floating point operands.</p>
3873<h5>Arguments:</h5>
3874<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
3875the condition code indicating the kind of comparison to perform. It is not
3876a value, just a keyword. The possible condition code are:
3877<ol>
3878 <li><tt>false</tt>: no comparison, always returns false</li>
3879 <li><tt>oeq</tt>: ordered and equal</li>
3880 <li><tt>ogt</tt>: ordered and greater than </li>
3881 <li><tt>oge</tt>: ordered and greater than or equal</li>
3882 <li><tt>olt</tt>: ordered and less than </li>
3883 <li><tt>ole</tt>: ordered and less than or equal</li>
3884 <li><tt>one</tt>: ordered and not equal</li>
3885 <li><tt>ord</tt>: ordered (no nans)</li>
3886 <li><tt>ueq</tt>: unordered or equal</li>
3887 <li><tt>ugt</tt>: unordered or greater than </li>
3888 <li><tt>uge</tt>: unordered or greater than or equal</li>
3889 <li><tt>ult</tt>: unordered or less than </li>
3890 <li><tt>ule</tt>: unordered or less than or equal</li>
3891 <li><tt>une</tt>: unordered or not equal</li>
3892 <li><tt>uno</tt>: unordered (either nans)</li>
3893 <li><tt>true</tt>: no comparison, always returns true</li>
3894</ol>
3895<p><i>Ordered</i> means that neither operand is a QNAN while
3896<i>unordered</i> means that either operand may be a QNAN.</p>
3897<p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3898<a href="#t_floating">floating point</a> typed. They must have identical
3899types.</p>
3900<h5>Semantics:</h5>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]3901<p>The '<tt>fcmp</tt>' instruction compares <tt>var1</tt> and <tt>var2</tt>
3902according to the condition code given as <tt>cond</tt>. The comparison performed
3903always yields a <a href="#t_primitive">i1</a> result, as follows:
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3904<ol>
3905 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
3906 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3907 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
3908 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3909 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
3910 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3911 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3912 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3913 <tt>var1</tt> is less than <tt>var2</tt>.</li>
3914 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3915 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3916 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3917 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
3918 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3919 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
3920 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
3921 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
3922 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3923 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
3924 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3925 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
3926 <tt>var1</tt> is less than <tt>var2</tt>.</li>
3927 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
3928 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3929 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
3930 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
3931 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
3932 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3933</ol>
3934
3935<h5>Example:</h5>
3936<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3937 &lt;result&gt; = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3938 &lt;result&gt; = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3939 &lt;result&gt; = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3940</pre>
3941</div>
3942
3943<!-- _______________________________________________________________________ -->
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]3944<div class="doc_subsubsection">
3945 <a name="i_vicmp">'<tt>vicmp</tt>' Instruction</a>
3946</div>
3947<div class="doc_text">
3948<h5>Syntax:</h5>
3949<pre> &lt;result&gt; = vicmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
3950</pre>
3951<h5>Overview:</h5>
3952<p>The '<tt>vicmp</tt>' instruction returns an integer vector value based on
3953element-wise comparison of its two integer vector operands.</p>
3954<h5>Arguments:</h5>
3955<p>The '<tt>vicmp</tt>' instruction takes three operands. The first operand is
3956the condition code indicating the kind of comparison to perform. It is not
3957a value, just a keyword. The possible condition code are:
3958<ol>
3959 <li><tt>eq</tt>: equal</li>
3960 <li><tt>ne</tt>: not equal </li>
3961 <li><tt>ugt</tt>: unsigned greater than</li>
3962 <li><tt>uge</tt>: unsigned greater or equal</li>
3963 <li><tt>ult</tt>: unsigned less than</li>
3964 <li><tt>ule</tt>: unsigned less or equal</li>
3965 <li><tt>sgt</tt>: signed greater than</li>
3966 <li><tt>sge</tt>: signed greater or equal</li>
3967 <li><tt>slt</tt>: signed less than</li>
3968 <li><tt>sle</tt>: signed less or equal</li>
3969</ol>
3970<p>The remaining two arguments must be <a href="#t_vector">vector</a> of
3971<a href="#t_integer">integer</a> typed. They must also be identical types.</p>
3972<h5>Semantics:</h5>
3973<p>The '<tt>vicmp</tt>' instruction compares <tt>var1</tt> and <tt>var2</tt>
3974according to the condition code given as <tt>cond</tt>. The comparison yields a
3975<a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, of
3976identical type as the values being compared. The most significant bit in each
3977element is 1 if the element-wise comparison evaluates to true, and is 0
3978otherwise. All other bits of the result are undefined. The condition codes
3979are evaluated identically to the <a href="#i_icmp">'<tt>icmp</tt>'
3980instruction</a>.
3981
3982<h5>Example:</h5>
3983<pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]3984 &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>
3985 &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]3986</pre>
3987</div>
3988
3989<!-- _______________________________________________________________________ -->
3990<div class="doc_subsubsection">
3991 <a name="i_vfcmp">'<tt>vfcmp</tt>' Instruction</a>
3992</div>
3993<div class="doc_text">
3994<h5>Syntax:</h5>
3995<pre> &lt;result&gt; = vfcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt;</pre>
3996<h5>Overview:</h5>
3997<p>The '<tt>vfcmp</tt>' instruction returns an integer vector value based on
3998element-wise comparison of its two floating point vector operands. The output
3999elements have the same width as the input elements.</p>
4000<h5>Arguments:</h5>
4001<p>The '<tt>vfcmp</tt>' instruction takes three operands. The first operand is
4002the condition code indicating the kind of comparison to perform. It is not
4003a value, just a keyword. The possible condition code are:
4004<ol>
4005 <li><tt>false</tt>: no comparison, always returns false</li>
4006 <li><tt>oeq</tt>: ordered and equal</li>
4007 <li><tt>ogt</tt>: ordered and greater than </li>
4008 <li><tt>oge</tt>: ordered and greater than or equal</li>
4009 <li><tt>olt</tt>: ordered and less than </li>
4010 <li><tt>ole</tt>: ordered and less than or equal</li>
4011 <li><tt>one</tt>: ordered and not equal</li>
4012 <li><tt>ord</tt>: ordered (no nans)</li>
4013 <li><tt>ueq</tt>: unordered or equal</li>
4014 <li><tt>ugt</tt>: unordered or greater than </li>
4015 <li><tt>uge</tt>: unordered or greater than or equal</li>
4016 <li><tt>ult</tt>: unordered or less than </li>
4017 <li><tt>ule</tt>: unordered or less than or equal</li>
4018 <li><tt>une</tt>: unordered or not equal</li>
4019 <li><tt>uno</tt>: unordered (either nans)</li>
4020 <li><tt>true</tt>: no comparison, always returns true</li>
4021</ol>
4022<p>The remaining two arguments must be <a href="#t_vector">vector</a> of
4023<a href="#t_floating">floating point</a> typed. They must also be identical
4024types.</p>
4025<h5>Semantics:</h5>
4026<p>The '<tt>vfcmp</tt>' instruction compares <tt>var1</tt> and <tt>var2</tt>
4027according to the condition code given as <tt>cond</tt>. The comparison yields a
4028<a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, with
4029an identical number of elements as the values being compared, and each element
4030having identical with to the width of the floating point elements. The most
4031significant bit in each element is 1 if the element-wise comparison evaluates to
4032true, and is 0 otherwise. All other bits of the result are undefined. The
4033condition codes are evaluated identically to the
4034<a href="#i_fcmp">'<tt>fcmp</tt>' instruction</a>.
4035
4036<h5>Example:</h5>
4037<pre>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4038 &lt;result&gt; = vfcmp oeq &lt;2 x float&gt; &lt; float 4, float 0 &gt;, &lt; float 5, float 0 &gt; <i>; yields: result=&lt;2 x i32&gt; &lt; i32 0, i32 -1 &gt;</i>
4039 &lt;result&gt; = vfcmp ult &lt;2 x double&gt; &lt; double 1, double 2 &gt;, &lt; double 2, double 2&gt; <i>; yields: result=&lt;2 x i64&gt; &lt; i64 -1, i64 0 &gt;</i>
Nate Begeman646fa482008-05-12 19:01:56 +0000[diff] [blame]4040</pre>
4041</div>
4042
4043<!-- _______________________________________________________________________ -->
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4044<div class="doc_subsubsection">
4045 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
4046</div>
4047
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4048<div class="doc_text">
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4049
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4050<h5>Syntax:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4051
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4052<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
4053<h5>Overview:</h5>
4054<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
4055the SSA graph representing the function.</p>
4056<h5>Arguments:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4057
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4058<p>The type of the incoming values is specified with the first type
4059field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
4060as arguments, with one pair for each predecessor basic block of the
4061current block. Only values of <a href="#t_firstclass">first class</a>
4062type may be used as the value arguments to the PHI node. Only labels
4063may be used as the label arguments.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4064
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4065<p>There must be no non-phi instructions between the start of a basic
4066block and the PHI instructions: i.e. PHI instructions must be first in
4067a basic block.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4068
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4069<h5>Semantics:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4070
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4071<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
4072specified by the pair corresponding to the predecessor basic block that executed
4073just prior to the current block.</p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4074
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4075<h5>Example:</h5>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4076<pre>
4077Loop: ; Infinite loop that counts from 0 on up...
4078 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
4079 %nextindvar = add i32 %indvar, 1
4080 br label %Loop
4081</pre>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4082</div>
4083
4084<!-- _______________________________________________________________________ -->
4085<div class="doc_subsubsection">
4086 <a name="i_select">'<tt>select</tt>' Instruction</a>
4087</div>
4088
4089<div class="doc_text">
4090
4091<h5>Syntax:</h5>
4092
4093<pre>
4094 &lt;result&gt; = select i1 &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
4095</pre>
4096
4097<h5>Overview:</h5>
4098
4099<p>
4100The '<tt>select</tt>' instruction is used to choose one value based on a
4101condition, without branching.
4102</p>
4103
4104
4105<h5>Arguments:</h5>
4106
4107<p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4108The '<tt>select</tt>' instruction requires an 'i1' value indicating the
4109condition, and two values of the same <a href="#t_firstclass">first class</a>
4110type. If the val1/val2 are vectors, the entire vectors are selected, not
4111individual elements.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4112</p>
4113
4114<h5>Semantics:</h5>
4115
4116<p>
Chris Lattner6704c212008-05-20 20:48:21 +0000[diff] [blame]4117If the i1 condition evaluates is 1, the instruction returns the first
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4118value argument; otherwise, it returns the second value argument.
4119</p>
4120
4121<h5>Example:</h5>
4122
4123<pre>
4124 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
4125</pre>
4126</div>
4127
4128
4129<!-- _______________________________________________________________________ -->
4130<div class="doc_subsubsection">
4131 <a name="i_call">'<tt>call</tt>' Instruction</a>
4132</div>
4133
4134<div class="doc_text">
4135
4136<h5>Syntax:</h5>
4137<pre>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]4138 &lt;result&gt; = [tail] call [<a href="#callingconv">cconv</a>] &lt;ty&gt; [&lt;fnty&gt;*] &lt;fnptrval&gt;(&lt;param list&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4139</pre>
4140
4141<h5>Overview:</h5>
4142
4143<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
4144
4145<h5>Arguments:</h5>
4146
4147<p>This instruction requires several arguments:</p>
4148
4149<ol>
4150 <li>
4151 <p>The optional "tail" marker indicates whether the callee function accesses
4152 any allocas or varargs in the caller. If the "tail" marker is present, the
4153 function call is eligible for tail call optimization. Note that calls may
4154 be marked "tail" even if they do not occur before a <a
4155 href="#i_ret"><tt>ret</tt></a> instruction.
4156 </li>
4157 <li>
4158 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
4159 convention</a> the call should use. If none is specified, the call defaults
4160 to using C calling conventions.
4161 </li>
4162 <li>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]4163 <p>'<tt>ty</tt>': the type of the call instruction itself which is also
4164 the type of the return value. Functions that return no value are marked
4165 <tt><a href="#t_void">void</a></tt>.</p>
4166 </li>
4167 <li>
4168 <p>'<tt>fnty</tt>': shall be the signature of the pointer to function
4169 value being invoked. The argument types must match the types implied by
4170 this signature. This type can be omitted if the function is not varargs
4171 and if the function type does not return a pointer to a function.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4172 </li>
4173 <li>
4174 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
4175 be invoked. In most cases, this is a direct function invocation, but
4176 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
4177 to function value.</p>
4178 </li>
4179 <li>
4180 <p>'<tt>function args</tt>': argument list whose types match the
4181 function signature argument types. All arguments must be of
4182 <a href="#t_firstclass">first class</a> type. If the function signature
4183 indicates the function accepts a variable number of arguments, the extra
4184 arguments can be specified.</p>
4185 </li>
4186</ol>
4187
4188<h5>Semantics:</h5>
4189
4190<p>The '<tt>call</tt>' instruction is used to cause control flow to
4191transfer to a specified function, with its incoming arguments bound to
4192the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
4193instruction in the called function, control flow continues with the
4194instruction after the function call, and the return value of the
Chris Lattner5e893ef2008-03-21 17:24:17 +0000[diff] [blame]4195function is bound to the result argument. If the callee returns multiple
4196values then the return values of the function are only accessible through
4197the '<tt><a href="#i_getresult">getresult</a></tt>' instruction.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4198
4199<h5>Example:</h5>
4200
4201<pre>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]4202 %retval = call i32 @test(i32 %argc)
Chris Lattner5e893ef2008-03-21 17:24:17 +0000[diff] [blame]4203 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42) <i>; yields i32</i>
4204 %X = tail call i32 @foo() <i>; yields i32</i>
4205 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
4206 call void %foo(i8 97 signext)
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4207
4208 %struct.A = type { i32, i8 }
Chris Lattner5e893ef2008-03-21 17:24:17 +0000[diff] [blame]4209 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
4210 %gr = getresult %struct.A %r, 0 <i>; yields i32</i>
4211 %gr1 = getresult %struct.A %r, 1 <i>; yields i8</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4212</pre>
4213
4214</div>
4215
4216<!-- _______________________________________________________________________ -->
4217<div class="doc_subsubsection">
4218 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
4219</div>
4220
4221<div class="doc_text">
4222
4223<h5>Syntax:</h5>
4224
4225<pre>
4226 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
4227</pre>
4228
4229<h5>Overview:</h5>
4230
4231<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
4232the "variable argument" area of a function call. It is used to implement the
4233<tt>va_arg</tt> macro in C.</p>
4234
4235<h5>Arguments:</h5>
4236
4237<p>This instruction takes a <tt>va_list*</tt> value and the type of
4238the argument. It returns a value of the specified argument type and
4239increments the <tt>va_list</tt> to point to the next argument. The
4240actual type of <tt>va_list</tt> is target specific.</p>
4241
4242<h5>Semantics:</h5>
4243
4244<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
4245type from the specified <tt>va_list</tt> and causes the
4246<tt>va_list</tt> to point to the next argument. For more information,
4247see the variable argument handling <a href="#int_varargs">Intrinsic
4248Functions</a>.</p>
4249
4250<p>It is legal for this instruction to be called in a function which does not
4251take a variable number of arguments, for example, the <tt>vfprintf</tt>
4252function.</p>
4253
4254<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
4255href="#intrinsics">intrinsic function</a> because it takes a type as an
4256argument.</p>
4257
4258<h5>Example:</h5>
4259
4260<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
4261
4262</div>
4263
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4264<!-- _______________________________________________________________________ -->
4265<div class="doc_subsubsection">
4266 <a name="i_getresult">'<tt>getresult</tt>' Instruction</a>
4267</div>
4268
4269<div class="doc_text">
4270
4271<h5>Syntax:</h5>
4272<pre>
Chris Lattneree9da3f2008-03-21 17:20:51 +0000[diff] [blame]4273 &lt;resultval&gt; = getresult &lt;type&gt; &lt;retval&gt;, &lt;index&gt;
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4274</pre>
Chris Lattneree9da3f2008-03-21 17:20:51 +0000[diff] [blame]4275
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4276<h5>Overview:</h5>
4277
4278<p> The '<tt>getresult</tt>' instruction is used to extract individual values
Chris Lattneree9da3f2008-03-21 17:20:51 +0000[diff] [blame]4279from a '<tt><a href="#i_call">call</a></tt>'
4280or '<tt><a href="#i_invoke">invoke</a></tt>' instruction that returns multiple
4281results.</p>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4282
4283<h5>Arguments:</h5>
4284
Chris Lattneree9da3f2008-03-21 17:20:51 +0000[diff] [blame]4285<p>The '<tt>getresult</tt>' instruction takes a call or invoke value as its
Chris Lattnerd8dd3522008-04-23 04:06:52 +0000[diff] [blame]4286first argument, or an undef value. The value must have <a
4287href="#t_struct">structure type</a>. The second argument is a constant
4288unsigned index value which must be in range for the number of values returned
4289by the call.</p>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4290
4291<h5>Semantics:</h5>
4292
Chris Lattneree9da3f2008-03-21 17:20:51 +0000[diff] [blame]4293<p>The '<tt>getresult</tt>' instruction extracts the element identified by
4294'<tt>index</tt>' from the aggregate value.</p>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4295
4296<h5>Example:</h5>
4297
4298<pre>
4299 %struct.A = type { i32, i8 }
4300
4301 %r = call %struct.A @foo()
Chris Lattneree9da3f2008-03-21 17:20:51 +0000[diff] [blame]4302 %gr = getresult %struct.A %r, 0 <i>; yields i32:%gr</i>
4303 %gr1 = getresult %struct.A %r, 1 <i>; yields i8:%gr1</i>
Devang Patela3cc5372008-03-10 20:49:15 +0000[diff] [blame]4304 add i32 %gr, 42
4305 add i8 %gr1, 41
4306</pre>
4307
4308</div>
4309
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4310<!-- *********************************************************************** -->
4311<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
4312<!-- *********************************************************************** -->
4313
4314<div class="doc_text">
4315
4316<p>LLVM supports the notion of an "intrinsic function". These functions have
4317well known names and semantics and are required to follow certain restrictions.
4318Overall, these intrinsics represent an extension mechanism for the LLVM
4319language that does not require changing all of the transformations in LLVM when
4320adding to the language (or the bitcode reader/writer, the parser, etc...).</p>
4321
4322<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
4323prefix is reserved in LLVM for intrinsic names; thus, function names may not
4324begin with this prefix. Intrinsic functions must always be external functions:
4325you cannot define the body of intrinsic functions. Intrinsic functions may
4326only be used in call or invoke instructions: it is illegal to take the address
4327of an intrinsic function. Additionally, because intrinsic functions are part
4328of the LLVM language, it is required if any are added that they be documented
4329here.</p>
4330
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4331<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
4332a family of functions that perform the same operation but on different data
4333types. Because LLVM can represent over 8 million different integer types,
4334overloading is used commonly to allow an intrinsic function to operate on any
4335integer type. One or more of the argument types or the result type can be
4336overloaded to accept any integer type. Argument types may also be defined as
4337exactly matching a previous argument's type or the result type. This allows an
4338intrinsic function which accepts multiple arguments, but needs all of them to
4339be of the same type, to only be overloaded with respect to a single argument or
4340the result.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4341
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4342<p>Overloaded intrinsics will have the names of its overloaded argument types
4343encoded into its function name, each preceded by a period. Only those types
4344which are overloaded result in a name suffix. Arguments whose type is matched
4345against another type do not. For example, the <tt>llvm.ctpop</tt> function can
4346take an integer of any width and returns an integer of exactly the same integer
4347width. This leads to a family of functions such as
4348<tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29 %val)</tt>.
4349Only one type, the return type, is overloaded, and only one type suffix is
4350required. Because the argument's type is matched against the return type, it
4351does not require its own name suffix.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4352
4353<p>To learn how to add an intrinsic function, please see the
4354<a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
4355</p>
4356
4357</div>
4358
4359<!-- ======================================================================= -->
4360<div class="doc_subsection">
4361 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
4362</div>
4363
4364<div class="doc_text">
4365
4366<p>Variable argument support is defined in LLVM with the <a
4367 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
4368intrinsic functions. These functions are related to the similarly
4369named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
4370
4371<p>All of these functions operate on arguments that use a
4372target-specific value type "<tt>va_list</tt>". The LLVM assembly
4373language reference manual does not define what this type is, so all
4374transformations should be prepared to handle these functions regardless of
4375the type used.</p>
4376
4377<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
4378instruction and the variable argument handling intrinsic functions are
4379used.</p>
4380
4381<div class="doc_code">
4382<pre>
4383define i32 @test(i32 %X, ...) {
4384 ; Initialize variable argument processing
4385 %ap = alloca i8*
4386 %ap2 = bitcast i8** %ap to i8*
4387 call void @llvm.va_start(i8* %ap2)
4388
4389 ; Read a single integer argument
4390 %tmp = va_arg i8** %ap, i32
4391
4392 ; Demonstrate usage of llvm.va_copy and llvm.va_end
4393 %aq = alloca i8*
4394 %aq2 = bitcast i8** %aq to i8*
4395 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
4396 call void @llvm.va_end(i8* %aq2)
4397
4398 ; Stop processing of arguments.
4399 call void @llvm.va_end(i8* %ap2)
4400 ret i32 %tmp
4401}
4402
4403declare void @llvm.va_start(i8*)
4404declare void @llvm.va_copy(i8*, i8*)
4405declare void @llvm.va_end(i8*)
4406</pre>
4407</div>
4408
4409</div>
4410
4411<!-- _______________________________________________________________________ -->
4412<div class="doc_subsubsection">
4413 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
4414</div>
4415
4416
4417<div class="doc_text">
4418<h5>Syntax:</h5>
4419<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
4420<h5>Overview:</h5>
4421<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
4422<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
4423href="#i_va_arg">va_arg</a></tt>.</p>
4424
4425<h5>Arguments:</h5>
4426
4427<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
4428
4429<h5>Semantics:</h5>
4430
4431<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
4432macro available in C. In a target-dependent way, it initializes the
4433<tt>va_list</tt> element to which the argument points, so that the next call to
4434<tt>va_arg</tt> will produce the first variable argument passed to the function.
4435Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
4436last argument of the function as the compiler can figure that out.</p>
4437
4438</div>
4439
4440<!-- _______________________________________________________________________ -->
4441<div class="doc_subsubsection">
4442 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
4443</div>
4444
4445<div class="doc_text">
4446<h5>Syntax:</h5>
4447<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
4448<h5>Overview:</h5>
4449
4450<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
4451which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
4452or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
4453
4454<h5>Arguments:</h5>
4455
4456<p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
4457
4458<h5>Semantics:</h5>
4459
4460<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
4461macro available in C. In a target-dependent way, it destroys the
4462<tt>va_list</tt> element to which the argument points. Calls to <a
4463href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
4464<tt>llvm.va_copy</tt></a> must be matched exactly with calls to
4465<tt>llvm.va_end</tt>.</p>
4466
4467</div>
4468
4469<!-- _______________________________________________________________________ -->
4470<div class="doc_subsubsection">
4471 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
4472</div>
4473
4474<div class="doc_text">
4475
4476<h5>Syntax:</h5>
4477
4478<pre>
4479 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
4480</pre>
4481
4482<h5>Overview:</h5>
4483
4484<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
4485from the source argument list to the destination argument list.</p>
4486
4487<h5>Arguments:</h5>
4488
4489<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
4490The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
4491
4492
4493<h5>Semantics:</h5>
4494
4495<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
4496macro available in C. In a target-dependent way, it copies the source
4497<tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
4498intrinsic is necessary because the <tt><a href="#int_va_start">
4499llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
4500example, memory allocation.</p>
4501
4502</div>
4503
4504<!-- ======================================================================= -->
4505<div class="doc_subsection">
4506 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
4507</div>
4508
4509<div class="doc_text">
4510
4511<p>
4512LLVM support for <a href="GarbageCollection.html">Accurate Garbage
4513Collection</a> requires the implementation and generation of these intrinsics.
4514These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
4515stack</a>, as well as garbage collector implementations that require <a
4516href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
4517Front-ends for type-safe garbage collected languages should generate these
4518intrinsics to make use of the LLVM garbage collectors. For more details, see <a
4519href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
4520</p>
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]4521
4522<p>The garbage collection intrinsics only operate on objects in the generic
4523 address space (address space zero).</p>
4524
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4525</div>
4526
4527<!-- _______________________________________________________________________ -->
4528<div class="doc_subsubsection">
4529 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
4530</div>
4531
4532<div class="doc_text">
4533
4534<h5>Syntax:</h5>
4535
4536<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4537 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4538</pre>
4539
4540<h5>Overview:</h5>
4541
4542<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
4543the code generator, and allows some metadata to be associated with it.</p>
4544
4545<h5>Arguments:</h5>
4546
4547<p>The first argument specifies the address of a stack object that contains the
4548root pointer. The second pointer (which must be either a constant or a global
4549value address) contains the meta-data to be associated with the root.</p>
4550
4551<h5>Semantics:</h5>
4552
Chris Lattnera7d94ba2008-04-24 05:59:56 +0000[diff] [blame]4553<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]4554location. At compile-time, the code generator generates information to allow
Gordon Henriksen40393542007-12-25 02:31:26 +0000[diff] [blame]4555the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
4556intrinsic may only be used in a function which <a href="#gc">specifies a GC
4557algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4558
4559</div>
4560
4561
4562<!-- _______________________________________________________________________ -->
4563<div class="doc_subsubsection">
4564 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
4565</div>
4566
4567<div class="doc_text">
4568
4569<h5>Syntax:</h5>
4570
4571<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4572 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4573</pre>
4574
4575<h5>Overview:</h5>
4576
4577<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
4578locations, allowing garbage collector implementations that require read
4579barriers.</p>
4580
4581<h5>Arguments:</h5>
4582
4583<p>The second argument is the address to read from, which should be an address
4584allocated from the garbage collector. The first object is a pointer to the
4585start of the referenced object, if needed by the language runtime (otherwise
4586null).</p>
4587
4588<h5>Semantics:</h5>
4589
4590<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
4591instruction, but may be replaced with substantially more complex code by the
Gordon Henriksen40393542007-12-25 02:31:26 +0000[diff] [blame]4592garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
4593may only be used in a function which <a href="#gc">specifies a GC
4594algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4595
4596</div>
4597
4598
4599<!-- _______________________________________________________________________ -->
4600<div class="doc_subsubsection">
4601 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
4602</div>
4603
4604<div class="doc_text">
4605
4606<h5>Syntax:</h5>
4607
4608<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4609 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4610</pre>
4611
4612<h5>Overview:</h5>
4613
4614<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
4615locations, allowing garbage collector implementations that require write
4616barriers (such as generational or reference counting collectors).</p>
4617
4618<h5>Arguments:</h5>
4619
4620<p>The first argument is the reference to store, the second is the start of the
4621object to store it to, and the third is the address of the field of Obj to
4622store to. If the runtime does not require a pointer to the object, Obj may be
4623null.</p>
4624
4625<h5>Semantics:</h5>
4626
4627<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
4628instruction, but may be replaced with substantially more complex code by the
Gordon Henriksen40393542007-12-25 02:31:26 +0000[diff] [blame]4629garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
4630may only be used in a function which <a href="#gc">specifies a GC
4631algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4632
4633</div>
4634
4635
4636
4637<!-- ======================================================================= -->
4638<div class="doc_subsection">
4639 <a name="int_codegen">Code Generator Intrinsics</a>
4640</div>
4641
4642<div class="doc_text">
4643<p>
4644These intrinsics are provided by LLVM to expose special features that may only
4645be implemented with code generator support.
4646</p>
4647
4648</div>
4649
4650<!-- _______________________________________________________________________ -->
4651<div class="doc_subsubsection">
4652 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
4653</div>
4654
4655<div class="doc_text">
4656
4657<h5>Syntax:</h5>
4658<pre>
4659 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
4660</pre>
4661
4662<h5>Overview:</h5>
4663
4664<p>
4665The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
4666target-specific value indicating the return address of the current function
4667or one of its callers.
4668</p>
4669
4670<h5>Arguments:</h5>
4671
4672<p>
4673The argument to this intrinsic indicates which function to return the address
4674for. Zero indicates the calling function, one indicates its caller, etc. The
4675argument is <b>required</b> to be a constant integer value.
4676</p>
4677
4678<h5>Semantics:</h5>
4679
4680<p>
4681The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
4682the return address of the specified call frame, or zero if it cannot be
4683identified. The value returned by this intrinsic is likely to be incorrect or 0
4684for arguments other than zero, so it should only be used for debugging purposes.
4685</p>
4686
4687<p>
4688Note that calling this intrinsic does not prevent function inlining or other
4689aggressive transformations, so the value returned may not be that of the obvious
4690source-language caller.
4691</p>
4692</div>
4693
4694
4695<!-- _______________________________________________________________________ -->
4696<div class="doc_subsubsection">
4697 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
4698</div>
4699
4700<div class="doc_text">
4701
4702<h5>Syntax:</h5>
4703<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4704 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4705</pre>
4706
4707<h5>Overview:</h5>
4708
4709<p>
4710The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
4711target-specific frame pointer value for the specified stack frame.
4712</p>
4713
4714<h5>Arguments:</h5>
4715
4716<p>
4717The argument to this intrinsic indicates which function to return the frame
4718pointer for. Zero indicates the calling function, one indicates its caller,
4719etc. The argument is <b>required</b> to be a constant integer value.
4720</p>
4721
4722<h5>Semantics:</h5>
4723
4724<p>
4725The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
4726the frame address of the specified call frame, or zero if it cannot be
4727identified. The value returned by this intrinsic is likely to be incorrect or 0
4728for arguments other than zero, so it should only be used for debugging purposes.
4729</p>
4730
4731<p>
4732Note that calling this intrinsic does not prevent function inlining or other
4733aggressive transformations, so the value returned may not be that of the obvious
4734source-language caller.
4735</p>
4736</div>
4737
4738<!-- _______________________________________________________________________ -->
4739<div class="doc_subsubsection">
4740 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
4741</div>
4742
4743<div class="doc_text">
4744
4745<h5>Syntax:</h5>
4746<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4747 declare i8 *@llvm.stacksave()
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4748</pre>
4749
4750<h5>Overview:</h5>
4751
4752<p>
4753The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
4754the function stack, for use with <a href="#int_stackrestore">
4755<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
4756features like scoped automatic variable sized arrays in C99.
4757</p>
4758
4759<h5>Semantics:</h5>
4760
4761<p>
4762This intrinsic returns a opaque pointer value that can be passed to <a
4763href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
4764<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
4765<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
4766state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
4767practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
4768that were allocated after the <tt>llvm.stacksave</tt> was executed.
4769</p>
4770
4771</div>
4772
4773<!-- _______________________________________________________________________ -->
4774<div class="doc_subsubsection">
4775 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
4776</div>
4777
4778<div class="doc_text">
4779
4780<h5>Syntax:</h5>
4781<pre>
4782 declare void @llvm.stackrestore(i8 * %ptr)
4783</pre>
4784
4785<h5>Overview:</h5>
4786
4787<p>
4788The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
4789the function stack to the state it was in when the corresponding <a
4790href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
4791useful for implementing language features like scoped automatic variable sized
4792arrays in C99.
4793</p>
4794
4795<h5>Semantics:</h5>
4796
4797<p>
4798See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
4799</p>
4800
4801</div>
4802
4803
4804<!-- _______________________________________________________________________ -->
4805<div class="doc_subsubsection">
4806 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
4807</div>
4808
4809<div class="doc_text">
4810
4811<h5>Syntax:</h5>
4812<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4813 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]4814</pre>
4815
4816<h5>Overview:</h5>
4817
4818
4819<p>
4820The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
4821a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
4822no
4823effect on the behavior of the program but can change its performance
4824characteristics.
4825</p>
4826
4827<h5>Arguments:</h5>
4828
4829<p>
4830<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
4831determining if the fetch should be for a read (0) or write (1), and
4832<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
4833locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
4834<tt>locality</tt> arguments must be constant integers.
4835</p>
4836
4837<h5>Semantics:</h5>
4838
4839<p>
4840This intrinsic does not modify the behavior of the program. In particular,
4841prefetches cannot trap and do not produce a value. On targets that support this
4842intrinsic, the prefetch can provide hints to the processor cache for better
4843performance.
4844</p>
4845
4846</div>
4847
4848<!-- _______________________________________________________________________ -->
4849<div class="doc_subsubsection">
4850 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
4851</div>
4852
4853<div class="doc_text">
4854
4855<h5>Syntax:</h5>
4856<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4857 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4858</pre>
4859
4860<h5>Overview:</h5>
4861
4862
4863<p>
4864The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
4865(PC) in a region of
4866code to simulators and other tools. The method is target specific, but it is
4867expected that the marker will use exported symbols to transmit the PC of the marker.
4868The marker makes no guarantees that it will remain with any specific instruction
4869after optimizations. It is possible that the presence of a marker will inhibit
4870optimizations. The intended use is to be inserted after optimizations to allow
4871correlations of simulation runs.
4872</p>
4873
4874<h5>Arguments:</h5>
4875
4876<p>
4877<tt>id</tt> is a numerical id identifying the marker.
4878</p>
4879
4880<h5>Semantics:</h5>
4881
4882<p>
4883This intrinsic does not modify the behavior of the program. Backends that do not
4884support this intrinisic may ignore it.
4885</p>
4886
4887</div>
4888
4889<!-- _______________________________________________________________________ -->
4890<div class="doc_subsubsection">
4891 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
4892</div>
4893
4894<div class="doc_text">
4895
4896<h5>Syntax:</h5>
4897<pre>
4898 declare i64 @llvm.readcyclecounter( )
4899</pre>
4900
4901<h5>Overview:</h5>
4902
4903
4904<p>
4905The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4906counter register (or similar low latency, high accuracy clocks) on those targets
4907that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4908As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4909should only be used for small timings.
4910</p>
4911
4912<h5>Semantics:</h5>
4913
4914<p>
4915When directly supported, reading the cycle counter should not modify any memory.
4916Implementations are allowed to either return a application specific value or a
4917system wide value. On backends without support, this is lowered to a constant 0.
4918</p>
4919
4920</div>
4921
4922<!-- ======================================================================= -->
4923<div class="doc_subsection">
4924 <a name="int_libc">Standard C Library Intrinsics</a>
4925</div>
4926
4927<div class="doc_text">
4928<p>
4929LLVM provides intrinsics for a few important standard C library functions.
4930These intrinsics allow source-language front-ends to pass information about the
4931alignment of the pointer arguments to the code generator, providing opportunity
4932for more efficient code generation.
4933</p>
4934
4935</div>
4936
4937<!-- _______________________________________________________________________ -->
4938<div class="doc_subsubsection">
4939 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
4940</div>
4941
4942<div class="doc_text">
4943
4944<h5>Syntax:</h5>
4945<pre>
4946 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4947 i32 &lt;len&gt;, i32 &lt;align&gt;)
4948 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4949 i64 &lt;len&gt;, i32 &lt;align&gt;)
4950</pre>
4951
4952<h5>Overview:</h5>
4953
4954<p>
4955The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
4956location to the destination location.
4957</p>
4958
4959<p>
4960Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4961intrinsics do not return a value, and takes an extra alignment argument.
4962</p>
4963
4964<h5>Arguments:</h5>
4965
4966<p>
4967The first argument is a pointer to the destination, the second is a pointer to
4968the source. The third argument is an integer argument
4969specifying the number of bytes to copy, and the fourth argument is the alignment
4970of the source and destination locations.
4971</p>
4972
4973<p>
4974If the call to this intrinisic has an alignment value that is not 0 or 1, then
4975the caller guarantees that both the source and destination pointers are aligned
4976to that boundary.
4977</p>
4978
4979<h5>Semantics:</h5>
4980
4981<p>
4982The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
4983location to the destination location, which are not allowed to overlap. It
4984copies "len" bytes of memory over. If the argument is known to be aligned to
4985some boundary, this can be specified as the fourth argument, otherwise it should
4986be set to 0 or 1.
4987</p>
4988</div>
4989
4990
4991<!-- _______________________________________________________________________ -->
4992<div class="doc_subsubsection">
4993 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
4994</div>
4995
4996<div class="doc_text">
4997
4998<h5>Syntax:</h5>
4999<pre>
5000 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5001 i32 &lt;len&gt;, i32 &lt;align&gt;)
5002 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
5003 i64 &lt;len&gt;, i32 &lt;align&gt;)
5004</pre>
5005
5006<h5>Overview:</h5>
5007
5008<p>
5009The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
5010location to the destination location. It is similar to the
Chris Lattnerdba16ea2008-01-06 19:51:52 +0000[diff] [blame]5011'<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to overlap.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5012</p>
5013
5014<p>
5015Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
5016intrinsics do not return a value, and takes an extra alignment argument.
5017</p>
5018
5019<h5>Arguments:</h5>
5020
5021<p>
5022The first argument is a pointer to the destination, the second is a pointer to
5023the source. The third argument is an integer argument
5024specifying the number of bytes to copy, and the fourth argument is the alignment
5025of the source and destination locations.
5026</p>
5027
5028<p>
5029If the call to this intrinisic has an alignment value that is not 0 or 1, then
5030the caller guarantees that the source and destination pointers are aligned to
5031that boundary.
5032</p>
5033
5034<h5>Semantics:</h5>
5035
5036<p>
5037The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
5038location to the destination location, which may overlap. It
5039copies "len" bytes of memory over. If the argument is known to be aligned to
5040some boundary, this can be specified as the fourth argument, otherwise it should
5041be set to 0 or 1.
5042</p>
5043</div>
5044
5045
5046<!-- _______________________________________________________________________ -->
5047<div class="doc_subsubsection">
5048 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
5049</div>
5050
5051<div class="doc_text">
5052
5053<h5>Syntax:</h5>
5054<pre>
5055 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5056 i32 &lt;len&gt;, i32 &lt;align&gt;)
5057 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
5058 i64 &lt;len&gt;, i32 &lt;align&gt;)
5059</pre>
5060
5061<h5>Overview:</h5>
5062
5063<p>
5064The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
5065byte value.
5066</p>
5067
5068<p>
5069Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
5070does not return a value, and takes an extra alignment argument.
5071</p>
5072
5073<h5>Arguments:</h5>
5074
5075<p>
5076The first argument is a pointer to the destination to fill, the second is the
5077byte value to fill it with, the third argument is an integer
5078argument specifying the number of bytes to fill, and the fourth argument is the
5079known alignment of destination location.
5080</p>
5081
5082<p>
5083If the call to this intrinisic has an alignment value that is not 0 or 1, then
5084the caller guarantees that the destination pointer is aligned to that boundary.
5085</p>
5086
5087<h5>Semantics:</h5>
5088
5089<p>
5090The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
5091the
5092destination location. If the argument is known to be aligned to some boundary,
5093this can be specified as the fourth argument, otherwise it should be set to 0 or
50941.
5095</p>
5096</div>
5097
5098
5099<!-- _______________________________________________________________________ -->
5100<div class="doc_subsubsection">
5101 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
5102</div>
5103
5104<div class="doc_text">
5105
5106<h5>Syntax:</h5>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5107<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]5108floating point or vector of floating point type. Not all targets support all
5109types however.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5110<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5111 declare float @llvm.sqrt.f32(float %Val)
5112 declare double @llvm.sqrt.f64(double %Val)
5113 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
5114 declare fp128 @llvm.sqrt.f128(fp128 %Val)
5115 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5116</pre>
5117
5118<h5>Overview:</h5>
5119
5120<p>
5121The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5122returning the same value as the libm '<tt>sqrt</tt>' functions would. Unlike
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5123<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
Chris Lattnerf914a002008-01-29 07:00:44 +0000[diff] [blame]5124negative numbers other than -0.0 (which allows for better optimization, because
5125there is no need to worry about errno being set). <tt>llvm.sqrt(-0.0)</tt> is
5126defined to return -0.0 like IEEE sqrt.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5127</p>
5128
5129<h5>Arguments:</h5>
5130
5131<p>
5132The argument and return value are floating point numbers of the same type.
5133</p>
5134
5135<h5>Semantics:</h5>
5136
5137<p>
5138This function returns the sqrt of the specified operand if it is a nonnegative
5139floating point number.
5140</p>
5141</div>
5142
5143<!-- _______________________________________________________________________ -->
5144<div class="doc_subsubsection">
5145 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
5146</div>
5147
5148<div class="doc_text">
5149
5150<h5>Syntax:</h5>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5151<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]5152floating point or vector of floating point type. Not all targets support all
5153types however.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5154<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]5155 declare float @llvm.powi.f32(float %Val, i32 %power)
5156 declare double @llvm.powi.f64(double %Val, i32 %power)
5157 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
5158 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
5159 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5160</pre>
5161
5162<h5>Overview:</h5>
5163
5164<p>
5165The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
5166specified (positive or negative) power. The order of evaluation of
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5167multiplications is not defined. When a vector of floating point type is
5168used, the second argument remains a scalar integer value.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5169</p>
5170
5171<h5>Arguments:</h5>
5172
5173<p>
5174The second argument is an integer power, and the first is a value to raise to
5175that power.
5176</p>
5177
5178<h5>Semantics:</h5>
5179
5180<p>
5181This function returns the first value raised to the second power with an
5182unspecified sequence of rounding operations.</p>
5183</div>
5184
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5185<!-- _______________________________________________________________________ -->
5186<div class="doc_subsubsection">
5187 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
5188</div>
5189
5190<div class="doc_text">
5191
5192<h5>Syntax:</h5>
5193<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
5194floating point or vector of floating point type. Not all targets support all
5195types however.
5196<pre>
5197 declare float @llvm.sin.f32(float %Val)
5198 declare double @llvm.sin.f64(double %Val)
5199 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
5200 declare fp128 @llvm.sin.f128(fp128 %Val)
5201 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
5202</pre>
5203
5204<h5>Overview:</h5>
5205
5206<p>
5207The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.
5208</p>
5209
5210<h5>Arguments:</h5>
5211
5212<p>
5213The argument and return value are floating point numbers of the same type.
5214</p>
5215
5216<h5>Semantics:</h5>
5217
5218<p>
5219This function returns the sine of the specified operand, returning the
5220same values as the libm <tt>sin</tt> functions would, and handles error
Dan Gohmaneaba92e2007-10-17 18:05:13 +0000[diff] [blame]5221conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5222</div>
5223
5224<!-- _______________________________________________________________________ -->
5225<div class="doc_subsubsection">
5226 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
5227</div>
5228
5229<div class="doc_text">
5230
5231<h5>Syntax:</h5>
5232<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
5233floating point or vector of floating point type. Not all targets support all
5234types however.
5235<pre>
5236 declare float @llvm.cos.f32(float %Val)
5237 declare double @llvm.cos.f64(double %Val)
5238 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
5239 declare fp128 @llvm.cos.f128(fp128 %Val)
5240 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
5241</pre>
5242
5243<h5>Overview:</h5>
5244
5245<p>
5246The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.
5247</p>
5248
5249<h5>Arguments:</h5>
5250
5251<p>
5252The argument and return value are floating point numbers of the same type.
5253</p>
5254
5255<h5>Semantics:</h5>
5256
5257<p>
5258This function returns the cosine of the specified operand, returning the
5259same values as the libm <tt>cos</tt> functions would, and handles error
Dan Gohmaneaba92e2007-10-17 18:05:13 +0000[diff] [blame]5260conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5261</div>
5262
5263<!-- _______________________________________________________________________ -->
5264<div class="doc_subsubsection">
5265 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
5266</div>
5267
5268<div class="doc_text">
5269
5270<h5>Syntax:</h5>
5271<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
5272floating point or vector of floating point type. Not all targets support all
5273types however.
5274<pre>
5275 declare float @llvm.pow.f32(float %Val, float %Power)
5276 declare double @llvm.pow.f64(double %Val, double %Power)
5277 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
5278 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
5279 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
5280</pre>
5281
5282<h5>Overview:</h5>
5283
5284<p>
5285The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
5286specified (positive or negative) power.
5287</p>
5288
5289<h5>Arguments:</h5>
5290
5291<p>
5292The second argument is a floating point power, and the first is a value to
5293raise to that power.
5294</p>
5295
5296<h5>Semantics:</h5>
5297
5298<p>
5299This function returns the first value raised to the second power,
5300returning the
5301same values as the libm <tt>pow</tt> functions would, and handles error
Dan Gohmaneaba92e2007-10-17 18:05:13 +0000[diff] [blame]5302conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]5303</div>
5304
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5305
5306<!-- ======================================================================= -->
5307<div class="doc_subsection">
5308 <a name="int_manip">Bit Manipulation Intrinsics</a>
5309</div>
5310
5311<div class="doc_text">
5312<p>
5313LLVM provides intrinsics for a few important bit manipulation operations.
5314These allow efficient code generation for some algorithms.
5315</p>
5316
5317</div>
5318
5319<!-- _______________________________________________________________________ -->
5320<div class="doc_subsubsection">
5321 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
5322</div>
5323
5324<div class="doc_text">
5325
5326<h5>Syntax:</h5>
5327<p>This is an overloaded intrinsic function. You can use bswap on any integer
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5328type that is an even number of bytes (i.e. BitWidth % 16 == 0).
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5329<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5330 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
5331 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
5332 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5333</pre>
5334
5335<h5>Overview:</h5>
5336
5337<p>
5338The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
5339values with an even number of bytes (positive multiple of 16 bits). These are
5340useful for performing operations on data that is not in the target's native
5341byte order.
5342</p>
5343
5344<h5>Semantics:</h5>
5345
5346<p>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5347The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5348and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
5349intrinsic returns an i32 value that has the four bytes of the input i32
5350swapped, 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]5351i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48</tt>,
5352<tt>llvm.bswap.i64</tt> and other intrinsics extend this concept to
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5353additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
5354</p>
5355
5356</div>
5357
5358<!-- _______________________________________________________________________ -->
5359<div class="doc_subsubsection">
5360 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
5361</div>
5362
5363<div class="doc_text">
5364
5365<h5>Syntax:</h5>
5366<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
5367width. Not all targets support all bit widths however.
5368<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5369 declare i8 @llvm.ctpop.i8 (i8 &lt;src&gt;)
5370 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5371 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5372 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
5373 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5374</pre>
5375
5376<h5>Overview:</h5>
5377
5378<p>
5379The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
5380value.
5381</p>
5382
5383<h5>Arguments:</h5>
5384
5385<p>
5386The only argument is the value to be counted. The argument may be of any
5387integer type. The return type must match the argument type.
5388</p>
5389
5390<h5>Semantics:</h5>
5391
5392<p>
5393The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
5394</p>
5395</div>
5396
5397<!-- _______________________________________________________________________ -->
5398<div class="doc_subsubsection">
5399 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
5400</div>
5401
5402<div class="doc_text">
5403
5404<h5>Syntax:</h5>
5405<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
5406integer bit width. Not all targets support all bit widths however.
5407<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5408 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
5409 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5410 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5411 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
5412 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5413</pre>
5414
5415<h5>Overview:</h5>
5416
5417<p>
5418The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
5419leading zeros in a variable.
5420</p>
5421
5422<h5>Arguments:</h5>
5423
5424<p>
5425The only argument is the value to be counted. The argument may be of any
5426integer type. The return type must match the argument type.
5427</p>
5428
5429<h5>Semantics:</h5>
5430
5431<p>
5432The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
5433in a variable. If the src == 0 then the result is the size in bits of the type
5434of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
5435</p>
5436</div>
5437
5438
5439
5440<!-- _______________________________________________________________________ -->
5441<div class="doc_subsubsection">
5442 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
5443</div>
5444
5445<div class="doc_text">
5446
5447<h5>Syntax:</h5>
5448<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
5449integer bit width. Not all targets support all bit widths however.
5450<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5451 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
5452 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5453 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5454 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
5455 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5456</pre>
5457
5458<h5>Overview:</h5>
5459
5460<p>
5461The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
5462trailing zeros.
5463</p>
5464
5465<h5>Arguments:</h5>
5466
5467<p>
5468The only argument is the value to be counted. The argument may be of any
5469integer type. The return type must match the argument type.
5470</p>
5471
5472<h5>Semantics:</h5>
5473
5474<p>
5475The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
5476in a variable. If the src == 0 then the result is the size in bits of the type
5477of src. For example, <tt>llvm.cttz(2) = 1</tt>.
5478</p>
5479</div>
5480
5481<!-- _______________________________________________________________________ -->
5482<div class="doc_subsubsection">
5483 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
5484</div>
5485
5486<div class="doc_text">
5487
5488<h5>Syntax:</h5>
5489<p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
5490on any integer bit width.
5491<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5492 declare i17 @llvm.part.select.i17 (i17 %val, i32 %loBit, i32 %hiBit)
5493 declare i29 @llvm.part.select.i29 (i29 %val, i32 %loBit, i32 %hiBit)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5494</pre>
5495
5496<h5>Overview:</h5>
5497<p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
5498range of bits from an integer value and returns them in the same bit width as
5499the original value.</p>
5500
5501<h5>Arguments:</h5>
5502<p>The first argument, <tt>%val</tt> and the result may be integer types of
5503any bit width but they must have the same bit width. The second and third
5504arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
5505
5506<h5>Semantics:</h5>
5507<p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
5508of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
5509<tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
5510operates in forward mode.</p>
5511<p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
5512right by <tt>%loBit</tt> bits and then ANDing it with a mask with
5513only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
5514<ol>
5515 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
5516 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
5517 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
5518 to determine the number of bits to retain.</li>
5519 <li>A mask of the retained bits is created by shifting a -1 value.</li>
5520 <li>The mask is ANDed with <tt>%val</tt> to produce the result.
5521</ol>
5522<p>In reverse mode, a similar computation is made except that the bits are
5523returned in the reverse order. So, for example, if <tt>X</tt> has the value
5524<tt>i16 0x0ACF (101011001111)</tt> and we apply
5525<tt>part.select(i16 X, 8, 3)</tt> to it, we get back the value
5526<tt>i16 0x0026 (000000100110)</tt>.</p>
5527</div>
5528
5529<div class="doc_subsubsection">
5530 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
5531</div>
5532
5533<div class="doc_text">
5534
5535<h5>Syntax:</h5>
5536<p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
5537on any integer bit width.
5538<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5539 declare i17 @llvm.part.set.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
5540 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]5541</pre>
5542
5543<h5>Overview:</h5>
5544<p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
5545of bits in an integer value with another integer value. It returns the integer
5546with the replaced bits.</p>
5547
5548<h5>Arguments:</h5>
5549<p>The first argument, <tt>%val</tt> and the result may be integer types of
5550any bit width but they must have the same bit width. <tt>%val</tt> is the value
5551whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
5552integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
5553type since they specify only a bit index.</p>
5554
5555<h5>Semantics:</h5>
5556<p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
5557of operation: forwards and reverse. If <tt>%lo</tt> is greater than
5558<tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
5559operates in forward mode.</p>
5560<p>For both modes, the <tt>%repl</tt> value is prepared for use by either
5561truncating it down to the size of the replacement area or zero extending it
5562up to that size.</p>
5563<p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
5564are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
5565in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
5566to the <tt>%hi</tt>th bit.
5567<p>In reverse mode, a similar computation is made except that the bits are
5568reversed. That is, the <tt>0</tt>th bit in <tt>%repl</tt> replaces the
5569<tt>%hi</tt> bit in <tt>%val</tt> and etc. down to the <tt>%lo</tt>th bit.
5570<h5>Examples:</h5>
5571<pre>
5572 llvm.part.set(0xFFFF, 0, 4, 7) -&gt; 0xFF0F
5573 llvm.part.set(0xFFFF, 0, 7, 4) -&gt; 0xFF0F
5574 llvm.part.set(0xFFFF, 1, 7, 4) -&gt; 0xFF8F
5575 llvm.part.set(0xFFFF, F, 8, 3) -&gt; 0xFFE7
5576 llvm.part.set(0xFFFF, 0, 3, 8) -&gt; 0xFE07
5577</pre>
5578</div>
5579
5580<!-- ======================================================================= -->
5581<div class="doc_subsection">
5582 <a name="int_debugger">Debugger Intrinsics</a>
5583</div>
5584
5585<div class="doc_text">
5586<p>
5587The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
5588are described in the <a
5589href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
5590Debugging</a> document.
5591</p>
5592</div>
5593
5594
5595<!-- ======================================================================= -->
5596<div class="doc_subsection">
5597 <a name="int_eh">Exception Handling Intrinsics</a>
5598</div>
5599
5600<div class="doc_text">
5601<p> The LLVM exception handling intrinsics (which all start with
5602<tt>llvm.eh.</tt> prefix), are described in the <a
5603href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
5604Handling</a> document. </p>
5605</div>
5606
5607<!-- ======================================================================= -->
5608<div class="doc_subsection">
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5609 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5610</div>
5611
5612<div class="doc_text">
5613<p>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5614 This intrinsic makes it possible to excise one parameter, marked with
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5615 the <tt>nest</tt> attribute, from a function. The result is a callable
5616 function pointer lacking the nest parameter - the caller does not need
5617 to provide a value for it. Instead, the value to use is stored in
5618 advance in a "trampoline", a block of memory usually allocated
5619 on the stack, which also contains code to splice the nest value into the
5620 argument list. This is used to implement the GCC nested function address
5621 extension.
5622</p>
5623<p>
5624 For example, if the function is
5625 <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]5626 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]5627<pre>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5628 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
5629 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
5630 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
5631 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5632</pre>
Bill Wendlinge262b4c2007-09-22 09:23:55 +0000[diff] [blame]5633 <p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
5634 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5635</div>
5636
5637<!-- _______________________________________________________________________ -->
5638<div class="doc_subsubsection">
5639 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
5640</div>
5641<div class="doc_text">
5642<h5>Syntax:</h5>
5643<pre>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5644declare 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]5645</pre>
5646<h5>Overview:</h5>
5647<p>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5648 This fills the memory pointed to by <tt>tramp</tt> with code
5649 and returns a function pointer suitable for executing it.
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5650</p>
5651<h5>Arguments:</h5>
5652<p>
5653 The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
5654 pointers. The <tt>tramp</tt> argument must point to a sufficiently large
5655 and sufficiently aligned block of memory; this memory is written to by the
Duncan Sands35012212007-08-22 23:39:54 +0000[diff] [blame]5656 intrinsic. Note that the size and the alignment are target-specific - LLVM
5657 currently provides no portable way of determining them, so a front-end that
5658 generates this intrinsic needs to have some target-specific knowledge.
5659 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]5660</p>
5661<h5>Semantics:</h5>
5662<p>
5663 The block of memory pointed to by <tt>tramp</tt> is filled with target
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5664 dependent code, turning it into a function. A pointer to this function is
5665 returned, but needs to be bitcast to an
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5666 <a href="#int_trampoline">appropriate function pointer type</a>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5667 before being called. The new function's signature is the same as that of
5668 <tt>func</tt> with any arguments marked with the <tt>nest</tt> attribute
5669 removed. At most one such <tt>nest</tt> argument is allowed, and it must be
5670 of pointer type. Calling the new function is equivalent to calling
5671 <tt>func</tt> with the same argument list, but with <tt>nval</tt> used for the
5672 missing <tt>nest</tt> argument. If, after calling
5673 <tt>llvm.init.trampoline</tt>, the memory pointed to by <tt>tramp</tt> is
5674 modified, then the effect of any later call to the returned function pointer is
5675 undefined.
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5676</p>
5677</div>
5678
5679<!-- ======================================================================= -->
5680<div class="doc_subsection">
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]5681 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
5682</div>
5683
5684<div class="doc_text">
5685<p>
5686 These intrinsic functions expand the "universal IR" of LLVM to represent
5687 hardware constructs for atomic operations and memory synchronization. This
5688 provides an interface to the hardware, not an interface to the programmer. It
5689 is aimed at a low enough level to allow any programming models or APIs which
5690 need atomic behaviors to map cleanly onto it. It is also modeled primarily on
5691 hardware behavior. Just as hardware provides a "universal IR" for source
5692 languages, it also provides a starting point for developing a "universal"
5693 atomic operation and synchronization IR.
5694</p>
5695<p>
5696 These do <em>not</em> form an API such as high-level threading libraries,
5697 software transaction memory systems, atomic primitives, and intrinsic
5698 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
5699 application libraries. The hardware interface provided by LLVM should allow
5700 a clean implementation of all of these APIs and parallel programming models.
5701 No one model or paradigm should be selected above others unless the hardware
5702 itself ubiquitously does so.
5703
5704</p>
5705</div>
5706
5707<!-- _______________________________________________________________________ -->
5708<div class="doc_subsubsection">
5709 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
5710</div>
5711<div class="doc_text">
5712<h5>Syntax:</h5>
5713<pre>
5714declare void @llvm.memory.barrier( i1 &lt;ll&gt;, i1 &lt;ls&gt;, i1 &lt;sl&gt;, i1 &lt;ss&gt;,
5715i1 &lt;device&gt; )
5716
5717</pre>
5718<h5>Overview:</h5>
5719<p>
5720 The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
5721 specific pairs of memory access types.
5722</p>
5723<h5>Arguments:</h5>
5724<p>
5725 The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
5726 The first four arguments enables a specific barrier as listed below. The fith
5727 argument specifies that the barrier applies to io or device or uncached memory.
5728
5729</p>
5730 <ul>
5731 <li><tt>ll</tt>: load-load barrier</li>
5732 <li><tt>ls</tt>: load-store barrier</li>
5733 <li><tt>sl</tt>: store-load barrier</li>
5734 <li><tt>ss</tt>: store-store barrier</li>
5735 <li><tt>device</tt>: barrier applies to device and uncached memory also.
5736 </ul>
5737<h5>Semantics:</h5>
5738<p>
5739 This intrinsic causes the system to enforce some ordering constraints upon
5740 the loads and stores of the program. This barrier does not indicate
5741 <em>when</em> any events will occur, it only enforces an <em>order</em> in
5742 which they occur. For any of the specified pairs of load and store operations
5743 (f.ex. load-load, or store-load), all of the first operations preceding the
5744 barrier will complete before any of the second operations succeeding the
5745 barrier begin. Specifically the semantics for each pairing is as follows:
5746</p>
5747 <ul>
5748 <li><tt>ll</tt>: All loads before the barrier must complete before any load
5749 after the barrier begins.</li>
5750
5751 <li><tt>ls</tt>: All loads before the barrier must complete before any
5752 store after the barrier begins.</li>
5753 <li><tt>ss</tt>: All stores before the barrier must complete before any
5754 store after the barrier begins.</li>
5755 <li><tt>sl</tt>: All stores before the barrier must complete before any
5756 load after the barrier begins.</li>
5757 </ul>
5758<p>
5759 These semantics are applied with a logical "and" behavior when more than one
5760 is enabled in a single memory barrier intrinsic.
5761</p>
5762<p>
5763 Backends may implement stronger barriers than those requested when they do not
5764 support as fine grained a barrier as requested. Some architectures do not
5765 need all types of barriers and on such architectures, these become noops.
5766</p>
5767<h5>Example:</h5>
5768<pre>
5769%ptr = malloc i32
5770 store i32 4, %ptr
5771
5772%result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
5773 call void @llvm.memory.barrier( i1 false, i1 true, i1 false, i1 false )
5774 <i>; guarantee the above finishes</i>
5775 store i32 8, %ptr <i>; before this begins</i>
5776</pre>
5777</div>
5778
Andrew Lenharthe44f3902008-02-21 06:45:13 +0000[diff] [blame]5779<!-- _______________________________________________________________________ -->
5780<div class="doc_subsubsection">
5781 <a name="int_atomic_lcs">'<tt>llvm.atomic.lcs.*</tt>' Intrinsic</a>
5782</div>
5783<div class="doc_text">
5784<h5>Syntax:</h5>
5785<p>
5786 This is an overloaded intrinsic. You can use <tt>llvm.atomic.lcs</tt> on any
5787 integer bit width. Not all targets support all bit widths however.</p>
5788
5789<pre>
5790declare i8 @llvm.atomic.lcs.i8( i8* &lt;ptr&gt;, i8 &lt;cmp&gt;, i8 &lt;val&gt; )
5791declare i16 @llvm.atomic.lcs.i16( i16* &lt;ptr&gt;, i16 &lt;cmp&gt;, i16 &lt;val&gt; )
5792declare i32 @llvm.atomic.lcs.i32( i32* &lt;ptr&gt;, i32 &lt;cmp&gt;, i32 &lt;val&gt; )
5793declare i64 @llvm.atomic.lcs.i64( i64* &lt;ptr&gt;, i64 &lt;cmp&gt;, i64 &lt;val&gt; )
5794
5795</pre>
5796<h5>Overview:</h5>
5797<p>
5798 This loads a value in memory and compares it to a given value. If they are
5799 equal, it stores a new value into the memory.
5800</p>
5801<h5>Arguments:</h5>
5802<p>
5803 The <tt>llvm.atomic.lcs</tt> intrinsic takes three arguments. The result as
5804 well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
5805 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
5806 this integer type. While any bit width integer may be used, targets may only
5807 lower representations they support in hardware.
5808
5809</p>
5810<h5>Semantics:</h5>
5811<p>
5812 This entire intrinsic must be executed atomically. It first loads the value
5813 in memory pointed to by <tt>ptr</tt> and compares it with the value
5814 <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the memory. The
5815 loaded value is yielded in all cases. This provides the equivalent of an
5816 atomic compare-and-swap operation within the SSA framework.
5817</p>
5818<h5>Examples:</h5>
5819
5820<pre>
5821%ptr = malloc i32
5822 store i32 4, %ptr
5823
5824%val1 = add i32 4, 4
5825%result1 = call i32 @llvm.atomic.lcs.i32( i32* %ptr, i32 4, %val1 )
5826 <i>; yields {i32}:result1 = 4</i>
5827%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
5828%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
5829
5830%val2 = add i32 1, 1
5831%result2 = call i32 @llvm.atomic.lcs.i32( i32* %ptr, i32 5, %val2 )
5832 <i>; yields {i32}:result2 = 8</i>
5833%stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
5834
5835%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
5836</pre>
5837</div>
5838
5839<!-- _______________________________________________________________________ -->
5840<div class="doc_subsubsection">
5841 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
5842</div>
5843<div class="doc_text">
5844<h5>Syntax:</h5>
5845
5846<p>
5847 This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
5848 integer bit width. Not all targets support all bit widths however.</p>
5849<pre>
5850declare i8 @llvm.atomic.swap.i8( i8* &lt;ptr&gt;, i8 &lt;val&gt; )
5851declare i16 @llvm.atomic.swap.i16( i16* &lt;ptr&gt;, i16 &lt;val&gt; )
5852declare i32 @llvm.atomic.swap.i32( i32* &lt;ptr&gt;, i32 &lt;val&gt; )
5853declare i64 @llvm.atomic.swap.i64( i64* &lt;ptr&gt;, i64 &lt;val&gt; )
5854
5855</pre>
5856<h5>Overview:</h5>
5857<p>
5858 This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
5859 the value from memory. It then stores the value in <tt>val</tt> in the memory
5860 at <tt>ptr</tt>.
5861</p>
5862<h5>Arguments:</h5>
5863
5864<p>
5865 The <tt>llvm.atomic.ls</tt> intrinsic takes two arguments. Both the
5866 <tt>val</tt> argument and the result must be integers of the same bit width.
5867 The first argument, <tt>ptr</tt>, must be a pointer to a value of this
5868 integer type. The targets may only lower integer representations they
5869 support.
5870</p>
5871<h5>Semantics:</h5>
5872<p>
5873 This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
5874 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
5875 equivalent of an atomic swap operation within the SSA framework.
5876
5877</p>
5878<h5>Examples:</h5>
5879<pre>
5880%ptr = malloc i32
5881 store i32 4, %ptr
5882
5883%val1 = add i32 4, 4
5884%result1 = call i32 @llvm.atomic.swap.i32( i32* %ptr, i32 %val1 )
5885 <i>; yields {i32}:result1 = 4</i>
5886%stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
5887%memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
5888
5889%val2 = add i32 1, 1
5890%result2 = call i32 @llvm.atomic.swap.i32( i32* %ptr, i32 %val2 )
5891 <i>; yields {i32}:result2 = 8</i>
5892
5893%stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
5894%memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
5895</pre>
5896</div>
5897
5898<!-- _______________________________________________________________________ -->
5899<div class="doc_subsubsection">
5900 <a name="int_atomic_las">'<tt>llvm.atomic.las.*</tt>' Intrinsic</a>
5901
5902</div>
5903<div class="doc_text">
5904<h5>Syntax:</h5>
5905<p>
5906 This is an overloaded intrinsic. You can use <tt>llvm.atomic.las</tt> on any
5907 integer bit width. Not all targets support all bit widths however.</p>
5908<pre>
5909declare i8 @llvm.atomic.las.i8.( i8* &lt;ptr&gt;, i8 &lt;delta&gt; )
5910declare i16 @llvm.atomic.las.i16.( i16* &lt;ptr&gt;, i16 &lt;delta&gt; )
5911declare i32 @llvm.atomic.las.i32.( i32* &lt;ptr&gt;, i32 &lt;delta&gt; )
5912declare i64 @llvm.atomic.las.i64.( i64* &lt;ptr&gt;, i64 &lt;delta&gt; )
5913
5914</pre>
5915<h5>Overview:</h5>
5916<p>
5917 This intrinsic adds <tt>delta</tt> to the value stored in memory at
5918 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.
5919</p>
5920<h5>Arguments:</h5>
5921<p>
5922
5923 The intrinsic takes two arguments, the first a pointer to an integer value
5924 and the second an integer value. The result is also an integer value. These
5925 integer types can have any bit width, but they must all have the same bit
5926 width. The targets may only lower integer representations they support.
5927</p>
5928<h5>Semantics:</h5>
5929<p>
5930 This intrinsic does a series of operations atomically. It first loads the
5931 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
5932 to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.
5933</p>
5934
5935<h5>Examples:</h5>
5936<pre>
5937%ptr = malloc i32
5938 store i32 4, %ptr
5939%result1 = call i32 @llvm.atomic.las.i32( i32* %ptr, i32 4 )
5940 <i>; yields {i32}:result1 = 4</i>
5941%result2 = call i32 @llvm.atomic.las.i32( i32* %ptr, i32 2 )
5942 <i>; yields {i32}:result2 = 8</i>
5943%result3 = call i32 @llvm.atomic.las.i32( i32* %ptr, i32 5 )
5944 <i>; yields {i32}:result3 = 10</i>
5945%memval = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
5946</pre>
5947</div>
5948
Andrew Lenharth785610d2008-02-16 01:24:58 +0000[diff] [blame]5949
5950<!-- ======================================================================= -->
5951<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5952 <a name="int_general">General Intrinsics</a>
5953</div>
5954
5955<div class="doc_text">
5956<p> This class of intrinsics is designed to be generic and has
5957no specific purpose. </p>
5958</div>
5959
5960<!-- _______________________________________________________________________ -->
5961<div class="doc_subsubsection">
5962 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
5963</div>
5964
5965<div class="doc_text">
5966
5967<h5>Syntax:</h5>
5968<pre>
5969 declare void @llvm.var.annotation(i8* &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5970</pre>
5971
5972<h5>Overview:</h5>
5973
5974<p>
5975The '<tt>llvm.var.annotation</tt>' intrinsic
5976</p>
5977
5978<h5>Arguments:</h5>
5979
5980<p>
5981The first argument is a pointer to a value, the second is a pointer to a
5982global string, the third is a pointer to a global string which is the source
5983file name, and the last argument is the line number.
5984</p>
5985
5986<h5>Semantics:</h5>
5987
5988<p>
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]5989This intrinsic allows annotation of local variables with arbitrary strings.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5990This can be useful for special purpose optimizations that want to look for these
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]5991annotations. These have no other defined use, they are ignored by code
5992generation and optimization.
5993</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5994</div>
5995
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]5996<!-- _______________________________________________________________________ -->
5997<div class="doc_subsubsection">
Tanya Lattnerc9869b12007-09-21 23:57:59 +0000[diff] [blame]5998 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]5999</div>
6000
6001<div class="doc_text">
6002
6003<h5>Syntax:</h5>
Tanya Lattnere545be72007-09-21 23:56:27 +0000[diff] [blame]6004<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
6005any integer bit width.
6006</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6007<pre>
Tanya Lattner09161fe2007-09-22 00:03:01 +0000[diff] [blame]6008 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6009 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6010 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6011 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
6012 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]6013</pre>
6014
6015<h5>Overview:</h5>
Tanya Lattnere545be72007-09-21 23:56:27 +0000[diff] [blame]6016
6017<p>
6018The '<tt>llvm.annotation</tt>' intrinsic.
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6019</p>
6020
6021<h5>Arguments:</h5>
6022
6023<p>
6024The first argument is an integer value (result of some expression),
6025the second is a pointer to a global string, the third is a pointer to a global
6026string which is the source file name, and the last argument is the line number.
Tanya Lattnere545be72007-09-21 23:56:27 +0000[diff] [blame]6027It returns the value of the first argument.
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]6028</p>
6029
6030<h5>Semantics:</h5>
6031
6032<p>
6033This intrinsic allows annotations to be put on arbitrary expressions
6034with arbitrary strings. This can be useful for special purpose optimizations
6035that want to look for these annotations. These have no other defined use, they
6036are ignored by code generation and optimization.
6037</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6038
Anton Korobeynikove6e764f2008-01-15 22:31:34 +0000[diff] [blame]6039<!-- _______________________________________________________________________ -->
6040<div class="doc_subsubsection">
6041 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
6042</div>
6043
6044<div class="doc_text">
6045
6046<h5>Syntax:</h5>
6047<pre>
6048 declare void @llvm.trap()
6049</pre>
6050
6051<h5>Overview:</h5>
6052
6053<p>
6054The '<tt>llvm.trap</tt>' intrinsic
6055</p>
6056
6057<h5>Arguments:</h5>
6058
6059<p>
6060None
6061</p>
6062
6063<h5>Semantics:</h5>
6064
6065<p>
6066This intrinsics is lowered to the target dependent trap instruction. If the
6067target does not have a trap instruction, this intrinsic will be lowered to the
6068call of the abort() function.
6069</p>
6070</div>
6071
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6072<!-- *********************************************************************** -->
6073<hr>
6074<address>
6075 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
6076 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
6077 <a href="http://validator.w3.org/check/referer"><img
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]6078 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6079
6080 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
6081 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
6082 Last modified: $Date$
6083</address>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]6084
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]6085</body>
6086</html>