Gitiles
Code Review Sign In
gerrit-public.fairphone.software / fp2-dev / platform / external / llvm / 8643761465be6f6e5f27b6aeaedb6611113b2ff4 / . / docs / LangRef.html
blob: 51c308c0a96ba841135361bf91d6b1412308331d [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>
114 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
115 <ol>
116 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
117 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
118 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
119 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
120 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
121 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
122 </ol>
123 </li>
124 <li><a href="#convertops">Conversion Operations</a>
125 <ol>
126 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
127 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
128 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
129 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
130 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
131 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
132 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
133 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
134 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
135 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
136 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
137 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
138 </ol>
139 <li><a href="#otherops">Other Operations</a>
140 <ol>
141 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
142 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
143 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
144 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
145 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
146 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
147 </ol>
148 </li>
149 </ol>
150 </li>
151 <li><a href="#intrinsics">Intrinsic Functions</a>
152 <ol>
153 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
154 <ol>
155 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
156 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
157 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
158 </ol>
159 </li>
160 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
161 <ol>
162 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
163 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
164 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
165 </ol>
166 </li>
167 <li><a href="#int_codegen">Code Generator Intrinsics</a>
168 <ol>
169 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
170 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
171 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
172 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
173 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
174 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
175 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
176 </ol>
177 </li>
178 <li><a href="#int_libc">Standard C Library Intrinsics</a>
179 <ol>
180 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
181 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
182 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
183 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
184 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]185 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
186 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
187 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]188 </ol>
189 </li>
190 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
191 <ol>
192 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
193 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
194 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
195 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
196 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
197 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
198 </ol>
199 </li>
200 <li><a href="#int_debugger">Debugger intrinsics</a></li>
201 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]202 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]203 <ol>
204 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]205 </ol>
206 </li>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]207 <li><a href="#int_general">General intrinsics</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]208 <ol>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]209 <li><a href="#int_var_annotation">
Tanya Lattner51369f32007-09-22 00:01:26 +0000[diff] [blame]210 <tt>llvm.var.annotation</tt>' Intrinsic</a></li>
Reid Spencerb043f672007-07-20 19:59:11 +0000[diff] [blame]211 </ol>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]212 <ol>
213 <li><a href="#int_annotation">
Tanya Lattner51369f32007-09-22 00:01:26 +0000[diff] [blame]214 <tt>llvm.annotation.*</tt>' Intrinsic</a></li>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]215 </ol>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]216 </li>
217 </ol>
218 </li>
219</ol>
220
221<div class="doc_author">
222 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
223 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
224</div>
225
226<!-- *********************************************************************** -->
227<div class="doc_section"> <a name="abstract">Abstract </a></div>
228<!-- *********************************************************************** -->
229
230<div class="doc_text">
231<p>This document is a reference manual for the LLVM assembly language.
232LLVM is an SSA based representation that provides type safety,
233low-level operations, flexibility, and the capability of representing
234'all' high-level languages cleanly. It is the common code
235representation used throughout all phases of the LLVM compilation
236strategy.</p>
237</div>
238
239<!-- *********************************************************************** -->
240<div class="doc_section"> <a name="introduction">Introduction</a> </div>
241<!-- *********************************************************************** -->
242
243<div class="doc_text">
244
245<p>The LLVM code representation is designed to be used in three
246different forms: as an in-memory compiler IR, as an on-disk bitcode
247representation (suitable for fast loading by a Just-In-Time compiler),
248and as a human readable assembly language representation. This allows
249LLVM to provide a powerful intermediate representation for efficient
250compiler transformations and analysis, while providing a natural means
251to debug and visualize the transformations. The three different forms
252of LLVM are all equivalent. This document describes the human readable
253representation and notation.</p>
254
255<p>The LLVM representation aims to be light-weight and low-level
256while being expressive, typed, and extensible at the same time. It
257aims to be a "universal IR" of sorts, by being at a low enough level
258that high-level ideas may be cleanly mapped to it (similar to how
259microprocessors are "universal IR's", allowing many source languages to
260be mapped to them). By providing type information, LLVM can be used as
261the target of optimizations: for example, through pointer analysis, it
262can be proven that a C automatic variable is never accessed outside of
263the current function... allowing it to be promoted to a simple SSA
264value instead of a memory location.</p>
265
266</div>
267
268<!-- _______________________________________________________________________ -->
269<div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
270
271<div class="doc_text">
272
273<p>It is important to note that this document describes 'well formed'
274LLVM assembly language. There is a difference between what the parser
275accepts and what is considered 'well formed'. For example, the
276following instruction is syntactically okay, but not well formed:</p>
277
278<div class="doc_code">
279<pre>
280%x = <a href="#i_add">add</a> i32 1, %x
281</pre>
282</div>
283
284<p>...because the definition of <tt>%x</tt> does not dominate all of
285its uses. The LLVM infrastructure provides a verification pass that may
286be used to verify that an LLVM module is well formed. This pass is
287automatically run by the parser after parsing input assembly and by
288the optimizer before it outputs bitcode. The violations pointed out
289by the verifier pass indicate bugs in transformation passes or input to
290the parser.</p>
291</div>
292
Chris Lattnera83fdc02007-10-03 17:34:29 +0000[diff] [blame]293<!-- Describe the typesetting conventions here. -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]294
295<!-- *********************************************************************** -->
296<div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
297<!-- *********************************************************************** -->
298
299<div class="doc_text">
300
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]301 <p>LLVM identifiers come in two basic types: global and local. Global
302 identifiers (functions, global variables) begin with the @ character. Local
303 identifiers (register names, types) begin with the % character. Additionally,
304 there are three different formats for identifiers, for different purposes:
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]305
306<ol>
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]307 <li>Named values are represented as a string of characters with their prefix.
308 For example, %foo, @DivisionByZero, %a.really.long.identifier. The actual
309 regular expression used is '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]310 Identifiers which require other characters in their names can be surrounded
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]311 with quotes. In this way, anything except a <tt>&quot;</tt> character can
312 be used in a named value.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]313
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]314 <li>Unnamed values are represented as an unsigned numeric value with their
315 prefix. For example, %12, @2, %44.</li>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]316
317 <li>Constants, which are described in a <a href="#constants">section about
318 constants</a>, below.</li>
319</ol>
320
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]321<p>LLVM requires that values start with a prefix for two reasons: Compilers
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]322don't need to worry about name clashes with reserved words, and the set of
323reserved words may be expanded in the future without penalty. Additionally,
324unnamed identifiers allow a compiler to quickly come up with a temporary
325variable without having to avoid symbol table conflicts.</p>
326
327<p>Reserved words in LLVM are very similar to reserved words in other
328languages. There are keywords for different opcodes
329('<tt><a href="#i_add">add</a></tt>',
330 '<tt><a href="#i_bitcast">bitcast</a></tt>',
331 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
332href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
333and others. These reserved words cannot conflict with variable names, because
Reid Spencerc8245b02007-08-07 14:34:28 +0000[diff] [blame]334none of them start with a prefix character ('%' or '@').</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]335
336<p>Here is an example of LLVM code to multiply the integer variable
337'<tt>%X</tt>' by 8:</p>
338
339<p>The easy way:</p>
340
341<div class="doc_code">
342<pre>
343%result = <a href="#i_mul">mul</a> i32 %X, 8
344</pre>
345</div>
346
347<p>After strength reduction:</p>
348
349<div class="doc_code">
350<pre>
351%result = <a href="#i_shl">shl</a> i32 %X, i8 3
352</pre>
353</div>
354
355<p>And the hard way:</p>
356
357<div class="doc_code">
358<pre>
359<a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
360<a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
361%result = <a href="#i_add">add</a> i32 %1, %1
362</pre>
363</div>
364
365<p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
366important lexical features of LLVM:</p>
367
368<ol>
369
370 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
371 line.</li>
372
373 <li>Unnamed temporaries are created when the result of a computation is not
374 assigned to a named value.</li>
375
376 <li>Unnamed temporaries are numbered sequentially</li>
377
378</ol>
379
380<p>...and it also shows a convention that we follow in this document. When
381demonstrating instructions, we will follow an instruction with a comment that
382defines the type and name of value produced. Comments are shown in italic
383text.</p>
384
385</div>
386
387<!-- *********************************************************************** -->
388<div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
389<!-- *********************************************************************** -->
390
391<!-- ======================================================================= -->
392<div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
393</div>
394
395<div class="doc_text">
396
397<p>LLVM programs are composed of "Module"s, each of which is a
398translation unit of the input programs. Each module consists of
399functions, global variables, and symbol table entries. Modules may be
400combined together with the LLVM linker, which merges function (and
401global variable) definitions, resolves forward declarations, and merges
402symbol table entries. Here is an example of the "hello world" module:</p>
403
404<div class="doc_code">
405<pre><i>; Declare the string constant as a global constant...</i>
406<a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a
407 href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>
408
409<i>; External declaration of the puts function</i>
410<a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
411
412<i>; Definition of main function</i>
413define i32 @main() { <i>; i32()* </i>
414 <i>; Convert [13x i8 ]* to i8 *...</i>
415 %cast210 = <a
416 href="#i_getelementptr">getelementptr</a> [13 x i8 ]* @.LC0, i64 0, i64 0 <i>; i8 *</i>
417
418 <i>; Call puts function to write out the string to stdout...</i>
419 <a
420 href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
421 <a
422 href="#i_ret">ret</a> i32 0<br>}<br>
423</pre>
424</div>
425
426<p>This example is made up of a <a href="#globalvars">global variable</a>
427named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
428function, and a <a href="#functionstructure">function definition</a>
429for "<tt>main</tt>".</p>
430
431<p>In general, a module is made up of a list of global values,
432where both functions and global variables are global values. Global values are
433represented by a pointer to a memory location (in this case, a pointer to an
434array of char, and a pointer to a function), and have one of the following <a
435href="#linkage">linkage types</a>.</p>
436
437</div>
438
439<!-- ======================================================================= -->
440<div class="doc_subsection">
441 <a name="linkage">Linkage Types</a>
442</div>
443
444<div class="doc_text">
445
446<p>
447All Global Variables and Functions have one of the following types of linkage:
448</p>
449
450<dl>
451
452 <dt><tt><b><a name="linkage_internal">internal</a></b></tt> </dt>
453
454 <dd>Global values with internal linkage are only directly accessible by
455 objects in the current module. In particular, linking code into a module with
456 an internal global value may cause the internal to be renamed as necessary to
457 avoid collisions. Because the symbol is internal to the module, all
458 references can be updated. This corresponds to the notion of the
459 '<tt>static</tt>' keyword in C.
460 </dd>
461
462 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
463
464 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
465 the same name when linkage occurs. This is typically used to implement
466 inline functions, templates, or other code which must be generated in each
467 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
468 allowed to be discarded.
469 </dd>
470
471 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
472
473 <dd>"<tt>weak</tt>" linkage is exactly the same as <tt>linkonce</tt> linkage,
474 except that unreferenced <tt>weak</tt> globals may not be discarded. This is
475 used for globals that may be emitted in multiple translation units, but that
476 are not guaranteed to be emitted into every translation unit that uses them.
477 One example of this are common globals in C, such as "<tt>int X;</tt>" at
478 global scope.
479 </dd>
480
481 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
482
483 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
484 pointer to array type. When two global variables with appending linkage are
485 linked together, the two global arrays are appended together. This is the
486 LLVM, typesafe, equivalent of having the system linker append together
487 "sections" with identical names when .o files are linked.
488 </dd>
489
490 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
491 <dd>The semantics of this linkage follow the ELF model: the symbol is weak
492 until linked, if not linked, the symbol becomes null instead of being an
493 undefined reference.
494 </dd>
495
496 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
497
498 <dd>If none of the above identifiers are used, the global is externally
499 visible, meaning that it participates in linkage and can be used to resolve
500 external symbol references.
501 </dd>
502</dl>
503
504 <p>
505 The next two types of linkage are targeted for Microsoft Windows platform
506 only. They are designed to support importing (exporting) symbols from (to)
507 DLLs.
508 </p>
509
510 <dl>
511 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
512
513 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
514 or variable via a global pointer to a pointer that is set up by the DLL
515 exporting the symbol. On Microsoft Windows targets, the pointer name is
516 formed by combining <code>_imp__</code> and the function or variable name.
517 </dd>
518
519 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
520
521 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
522 pointer to a pointer in a DLL, so that it can be referenced with the
523 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
524 name is formed by combining <code>_imp__</code> and the function or variable
525 name.
526 </dd>
527
528</dl>
529
530<p><a name="linkage_external"></a>For example, since the "<tt>.LC0</tt>"
531variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
532variable and was linked with this one, one of the two would be renamed,
533preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
534external (i.e., lacking any linkage declarations), they are accessible
535outside of the current module.</p>
536<p>It is illegal for a function <i>declaration</i>
537to have any linkage type other than "externally visible", <tt>dllimport</tt>,
538or <tt>extern_weak</tt>.</p>
539<p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
540linkages.
541</div>
542
543<!-- ======================================================================= -->
544<div class="doc_subsection">
545 <a name="callingconv">Calling Conventions</a>
546</div>
547
548<div class="doc_text">
549
550<p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
551and <a href="#i_invoke">invokes</a> can all have an optional calling convention
552specified for the call. The calling convention of any pair of dynamic
553caller/callee must match, or the behavior of the program is undefined. The
554following calling conventions are supported by LLVM, and more may be added in
555the future:</p>
556
557<dl>
558 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
559
560 <dd>This calling convention (the default if no other calling convention is
561 specified) matches the target C calling conventions. This calling convention
562 supports varargs function calls and tolerates some mismatch in the declared
563 prototype and implemented declaration of the function (as does normal C).
564 </dd>
565
566 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
567
568 <dd>This calling convention attempts to make calls as fast as possible
569 (e.g. by passing things in registers). This calling convention allows the
570 target to use whatever tricks it wants to produce fast code for the target,
571 without having to conform to an externally specified ABI. Implementations of
572 this convention should allow arbitrary tail call optimization to be supported.
573 This calling convention does not support varargs and requires the prototype of
574 all callees to exactly match the prototype of the function definition.
575 </dd>
576
577 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
578
579 <dd>This calling convention attempts to make code in the caller as efficient
580 as possible under the assumption that the call is not commonly executed. As
581 such, these calls often preserve all registers so that the call does not break
582 any live ranges in the caller side. This calling convention does not support
583 varargs and requires the prototype of all callees to exactly match the
584 prototype of the function definition.
585 </dd>
586
587 <dt><b>"<tt>cc &lt;<em>n</em>&gt;</tt>" - Numbered convention</b>:</dt>
588
589 <dd>Any calling convention may be specified by number, allowing
590 target-specific calling conventions to be used. Target specific calling
591 conventions start at 64.
592 </dd>
593</dl>
594
595<p>More calling conventions can be added/defined on an as-needed basis, to
596support pascal conventions or any other well-known target-independent
597convention.</p>
598
599</div>
600
601<!-- ======================================================================= -->
602<div class="doc_subsection">
603 <a name="visibility">Visibility Styles</a>
604</div>
605
606<div class="doc_text">
607
608<p>
609All Global Variables and Functions have one of the following visibility styles:
610</p>
611
612<dl>
613 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
614
615 <dd>On ELF, default visibility means that the declaration is visible to other
616 modules and, in shared libraries, means that the declared entity may be
617 overridden. On Darwin, default visibility means that the declaration is
618 visible to other modules. Default visibility corresponds to "external
619 linkage" in the language.
620 </dd>
621
622 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
623
624 <dd>Two declarations of an object with hidden visibility refer to the same
625 object if they are in the same shared object. Usually, hidden visibility
626 indicates that the symbol will not be placed into the dynamic symbol table,
627 so no other module (executable or shared library) can reference it
628 directly.
629 </dd>
630
631 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
632
633 <dd>On ELF, protected visibility indicates that the symbol will be placed in
634 the dynamic symbol table, but that references within the defining module will
635 bind to the local symbol. That is, the symbol cannot be overridden by another
636 module.
637 </dd>
638</dl>
639
640</div>
641
642<!-- ======================================================================= -->
643<div class="doc_subsection">
644 <a name="globalvars">Global Variables</a>
645</div>
646
647<div class="doc_text">
648
649<p>Global variables define regions of memory allocated at compilation time
650instead of run-time. Global variables may optionally be initialized, may have
651an explicit section to be placed in, and may have an optional explicit alignment
652specified. A variable may be defined as "thread_local", which means that it
653will not be shared by threads (each thread will have a separated copy of the
654variable). A variable may be defined as a global "constant," which indicates
655that the contents of the variable will <b>never</b> be modified (enabling better
656optimization, allowing the global data to be placed in the read-only section of
657an executable, etc). Note that variables that need runtime initialization
658cannot be marked "constant" as there is a store to the variable.</p>
659
660<p>
661LLVM explicitly allows <em>declarations</em> of global variables to be marked
662constant, even if the final definition of the global is not. This capability
663can be used to enable slightly better optimization of the program, but requires
664the language definition to guarantee that optimizations based on the
665'constantness' are valid for the translation units that do not include the
666definition.
667</p>
668
669<p>As SSA values, global variables define pointer values that are in
670scope (i.e. they dominate) all basic blocks in the program. Global
671variables always define a pointer to their "content" type because they
672describe a region of memory, and all memory objects in LLVM are
673accessed through pointers.</p>
674
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]675<p>A global variable may be declared to reside in a target-specifc numbered
676address space. For targets that support them, address spaces may affect how
677optimizations are performed and/or what target instructions are used to access
Christopher Lamb20a39e92007-12-12 08:44:39 +0000[diff] [blame]678the variable. The default address space is zero. The address space qualifier
679must precede any other attributes.</p>
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]680
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]681<p>LLVM allows an explicit section to be specified for globals. If the target
682supports it, it will emit globals to the section specified.</p>
683
684<p>An explicit alignment may be specified for a global. If not present, or if
685the alignment is set to zero, the alignment of the global is set by the target
686to whatever it feels convenient. If an explicit alignment is specified, the
687global is forced to have at least that much alignment. All alignments must be
688a power of 2.</p>
689
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]690<p>For example, the following defines a global in a numbered address space with
691an initializer, section, and alignment:</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]692
693<div class="doc_code">
694<pre>
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]695@G = constant float 1.0 addrspace(5), section "foo", align 4
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]696</pre>
697</div>
698
699</div>
700
701
702<!-- ======================================================================= -->
703<div class="doc_subsection">
704 <a name="functionstructure">Functions</a>
705</div>
706
707<div class="doc_text">
708
709<p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
710an optional <a href="#linkage">linkage type</a>, an optional
711<a href="#visibility">visibility style</a>, an optional
712<a href="#callingconv">calling convention</a>, a return type, an optional
713<a href="#paramattrs">parameter attribute</a> for the return type, a function
714name, a (possibly empty) argument list (each with optional
715<a href="#paramattrs">parameter attributes</a>), an optional section, an
Gordon Henriksend606f5b2007-12-10 03:30:21 +0000[diff] [blame]716optional alignment, an optional <a href="#gc">garbage collector name</a>, an
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]717opening curly brace, a list of basic blocks, and a closing curly brace.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]718
719LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
720optional <a href="#linkage">linkage type</a>, an optional
721<a href="#visibility">visibility style</a>, an optional
722<a href="#callingconv">calling convention</a>, a return type, an optional
723<a href="#paramattrs">parameter attribute</a> for the return type, a function
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]724name, a possibly empty list of arguments, an optional alignment, and an optional
Gordon Henriksend606f5b2007-12-10 03:30:21 +0000[diff] [blame]725<a href="#gc">garbage collector name</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]726
727<p>A function definition contains a list of basic blocks, forming the CFG for
728the function. Each basic block may optionally start with a label (giving the
729basic block a symbol table entry), contains a list of instructions, and ends
730with a <a href="#terminators">terminator</a> instruction (such as a branch or
731function return).</p>
732
733<p>The first basic block in a function is special in two ways: it is immediately
734executed on entrance to the function, and it is not allowed to have predecessor
735basic blocks (i.e. there can not be any branches to the entry block of a
736function). Because the block can have no predecessors, it also cannot have any
737<a href="#i_phi">PHI nodes</a>.</p>
738
739<p>LLVM allows an explicit section to be specified for functions. If the target
740supports it, it will emit functions to the section specified.</p>
741
742<p>An explicit alignment may be specified for a function. If not present, or if
743the alignment is set to zero, the alignment of the function is set by the target
744to whatever it feels convenient. If an explicit alignment is specified, the
745function is forced to have at least that much alignment. All alignments must be
746a power of 2.</p>
747
748</div>
749
750
751<!-- ======================================================================= -->
752<div class="doc_subsection">
753 <a name="aliasstructure">Aliases</a>
754</div>
755<div class="doc_text">
756 <p>Aliases act as "second name" for the aliasee value (which can be either
757 function or global variable or bitcast of global value). Aliases may have an
758 optional <a href="#linkage">linkage type</a>, and an
759 optional <a href="#visibility">visibility style</a>.</p>
760
761 <h5>Syntax:</h5>
762
763<div class="doc_code">
764<pre>
765@&lt;Name&gt; = [Linkage] [Visibility] alias &lt;AliaseeTy&gt; @&lt;Aliasee&gt;
766</pre>
767</div>
768
769</div>
770
771
772
773<!-- ======================================================================= -->
774<div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
775<div class="doc_text">
776 <p>The return type and each parameter of a function type may have a set of
777 <i>parameter attributes</i> associated with them. Parameter attributes are
778 used to communicate additional information about the result or parameters of
Duncan Sandsf5588dc2007-11-27 13:23:08 +0000[diff] [blame]779 a function. Parameter attributes are considered to be part of the function,
780 not of the function type, so functions with different parameter attributes
781 can have the same function type.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]782
783 <p>Parameter attributes are simple keywords that follow the type specified. If
784 multiple parameter attributes are needed, they are space separated. For
785 example:</p>
786
787<div class="doc_code">
788<pre>
Duncan Sandsf5588dc2007-11-27 13:23:08 +0000[diff] [blame]789declare i32 @printf(i8* noalias , ...) nounwind
790declare i32 @atoi(i8*) nounwind readonly
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]791</pre>
792</div>
793
Duncan Sandsf5588dc2007-11-27 13:23:08 +0000[diff] [blame]794 <p>Note that any attributes for the function result (<tt>nounwind</tt>,
795 <tt>readonly</tt>) come immediately after the argument list.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]796
797 <p>Currently, only the following parameter attributes are defined:</p>
798 <dl>
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]799 <dt><tt>zeroext</tt></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]800 <dd>This indicates that the parameter should be zero extended just before
801 a call to this function.</dd>
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]802 <dt><tt>signext</tt></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]803 <dd>This indicates that the parameter should be sign extended just before
804 a call to this function.</dd>
805 <dt><tt>inreg</tt></dt>
806 <dd>This indicates that the parameter should be placed in register (if
807 possible) during assembling function call. Support for this attribute is
808 target-specific</dd>
809 <dt><tt>sret</tt></dt>
810 <dd>This indicates that the parameter specifies the address of a structure
811 that is the return value of the function in the source program.</dd>
812 <dt><tt>noalias</tt></dt>
813 <dd>This indicates that the parameter not alias any other object or any
814 other "noalias" objects during the function call.
815 <dt><tt>noreturn</tt></dt>
816 <dd>This function attribute indicates that the function never returns. This
817 indicates to LLVM that every call to this function should be treated as if
818 an <tt>unreachable</tt> instruction immediately followed the call.</dd>
819 <dt><tt>nounwind</tt></dt>
820 <dd>This function attribute indicates that the function type does not use
821 the unwind instruction and does not allow stack unwinding to propagate
822 through it.</dd>
Duncan Sands4ee46812007-07-27 19:57:41 +0000[diff] [blame]823 <dt><tt>nest</tt></dt>
824 <dd>This indicates that the parameter can be excised using the
825 <a href="#int_trampoline">trampoline intrinsics</a>.</dd>
Duncan Sands13e13f82007-11-22 20:23:04 +0000[diff] [blame]826 <dt><tt>readonly</tt></dt>
Duncan Sandsd69c0e62007-11-14 21:14:02 +0000[diff] [blame]827 <dd>This function attribute indicates that the function has no side-effects
Duncan Sands13e13f82007-11-22 20:23:04 +0000[diff] [blame]828 except for producing a return value or throwing an exception. The value
829 returned must only depend on the function arguments and/or global variables.
830 It may use values obtained by dereferencing pointers.</dd>
831 <dt><tt>readnone</tt></dt>
832 <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]833 function, but in addition it is not allowed to dereference any pointer arguments
834 or global variables.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]835 </dl>
836
837</div>
838
839<!-- ======================================================================= -->
840<div class="doc_subsection">
Gordon Henriksen13fe5e32007-12-10 03:18:06 +0000[diff] [blame]841 <a name="gc">Garbage Collector Names</a>
842</div>
843
844<div class="doc_text">
845<p>Each function may specify a garbage collector name, which is simply a
846string.</p>
847
848<div class="doc_code"><pre
849>define void @f() gc "name" { ...</pre></div>
850
851<p>The compiler declares the supported values of <i>name</i>. Specifying a
852collector which will cause the compiler to alter its output in order to support
853the named garbage collection algorithm.</p>
854</div>
855
856<!-- ======================================================================= -->
857<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]858 <a name="moduleasm">Module-Level Inline Assembly</a>
859</div>
860
861<div class="doc_text">
862<p>
863Modules may contain "module-level inline asm" blocks, which corresponds to the
864GCC "file scope inline asm" blocks. These blocks are internally concatenated by
865LLVM and treated as a single unit, but may be separated in the .ll file if
866desired. The syntax is very simple:
867</p>
868
869<div class="doc_code">
870<pre>
871module asm "inline asm code goes here"
872module asm "more can go here"
873</pre>
874</div>
875
876<p>The strings can contain any character by escaping non-printable characters.
877 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
878 for the number.
879</p>
880
881<p>
882 The inline asm code is simply printed to the machine code .s file when
883 assembly code is generated.
884</p>
885</div>
886
887<!-- ======================================================================= -->
888<div class="doc_subsection">
889 <a name="datalayout">Data Layout</a>
890</div>
891
892<div class="doc_text">
893<p>A module may specify a target specific data layout string that specifies how
894data is to be laid out in memory. The syntax for the data layout is simply:</p>
895<pre> target datalayout = "<i>layout specification</i>"</pre>
896<p>The <i>layout specification</i> consists of a list of specifications
897separated by the minus sign character ('-'). Each specification starts with a
898letter and may include other information after the letter to define some
899aspect of the data layout. The specifications accepted are as follows: </p>
900<dl>
901 <dt><tt>E</tt></dt>
902 <dd>Specifies that the target lays out data in big-endian form. That is, the
903 bits with the most significance have the lowest address location.</dd>
904 <dt><tt>e</tt></dt>
905 <dd>Specifies that hte target lays out data in little-endian form. That is,
906 the bits with the least significance have the lowest address location.</dd>
907 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
908 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
909 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
910 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
911 too.</dd>
912 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
913 <dd>This specifies the alignment for an integer type of a given bit
914 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
915 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
916 <dd>This specifies the alignment for a vector type of a given bit
917 <i>size</i>.</dd>
918 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
919 <dd>This specifies the alignment for a floating point type of a given bit
920 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
921 (double).</dd>
922 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
923 <dd>This specifies the alignment for an aggregate type of a given bit
924 <i>size</i>.</dd>
925</dl>
926<p>When constructing the data layout for a given target, LLVM starts with a
927default set of specifications which are then (possibly) overriden by the
928specifications in the <tt>datalayout</tt> keyword. The default specifications
929are given in this list:</p>
930<ul>
931 <li><tt>E</tt> - big endian</li>
932 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
933 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
934 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
935 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
936 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
937 <li><tt>i64:32:64</tt> - i64 has abi alignment of 32-bits but preferred
938 alignment of 64-bits</li>
939 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
940 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
941 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
942 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
943 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
944</ul>
945<p>When llvm is determining the alignment for a given type, it uses the
946following rules:
947<ol>
948 <li>If the type sought is an exact match for one of the specifications, that
949 specification is used.</li>
950 <li>If no match is found, and the type sought is an integer type, then the
951 smallest integer type that is larger than the bitwidth of the sought type is
952 used. If none of the specifications are larger than the bitwidth then the the
953 largest integer type is used. For example, given the default specifications
954 above, the i7 type will use the alignment of i8 (next largest) while both
955 i65 and i256 will use the alignment of i64 (largest specified).</li>
956 <li>If no match is found, and the type sought is a vector type, then the
957 largest vector type that is smaller than the sought vector type will be used
958 as a fall back. This happens because <128 x double> can be implemented in
959 terms of 64 <2 x double>, for example.</li>
960</ol>
961</div>
962
963<!-- *********************************************************************** -->
964<div class="doc_section"> <a name="typesystem">Type System</a> </div>
965<!-- *********************************************************************** -->
966
967<div class="doc_text">
968
969<p>The LLVM type system is one of the most important features of the
970intermediate representation. Being typed enables a number of
971optimizations to be performed on the IR directly, without having to do
972extra analyses on the side before the transformation. A strong type
973system makes it easier to read the generated code and enables novel
974analyses and transformations that are not feasible to perform on normal
975three address code representations.</p>
976
977</div>
978
979<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]980<div class="doc_subsection"> <a name="t_classifications">Type
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]981Classifications</a> </div>
982<div class="doc_text">
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]983<p>The types fall into a few useful
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]984classifications:</p>
985
986<table border="1" cellspacing="0" cellpadding="4">
987 <tbody>
988 <tr><th>Classification</th><th>Types</th></tr>
989 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]990 <td><a href="#t_integer">integer</a></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]991 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
992 </tr>
993 <tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]994 <td><a href="#t_floating">floating point</a></td>
995 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]996 </tr>
997 <tr>
998 <td><a name="t_firstclass">first class</a></td>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]999 <td><a href="#t_integer">integer</a>,
1000 <a href="#t_floating">floating point</a>,
1001 <a href="#t_pointer">pointer</a>,
1002 <a href="#t_vector">vector</a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1003 </td>
1004 </tr>
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1005 <tr>
1006 <td><a href="#t_primitive">primitive</a></td>
1007 <td><a href="#t_label">label</a>,
1008 <a href="#t_void">void</a>,
1009 <a href="#t_integer">integer</a>,
1010 <a href="#t_floating">floating point</a>.</td>
1011 </tr>
1012 <tr>
1013 <td><a href="#t_derived">derived</a></td>
1014 <td><a href="#t_integer">integer</a>,
1015 <a href="#t_array">array</a>,
1016 <a href="#t_function">function</a>,
1017 <a href="#t_pointer">pointer</a>,
1018 <a href="#t_struct">structure</a>,
1019 <a href="#t_pstruct">packed structure</a>,
1020 <a href="#t_vector">vector</a>,
1021 <a href="#t_opaque">opaque</a>.
1022 </tr>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1023 </tbody>
1024</table>
1025
1026<p>The <a href="#t_firstclass">first class</a> types are perhaps the
1027most important. Values of these types are the only ones which can be
1028produced by instructions, passed as arguments, or used as operands to
1029instructions. This means that all structures and arrays must be
1030manipulated either by pointer or by component.</p>
1031</div>
1032
1033<!-- ======================================================================= -->
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1034<div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame^]1035
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1036<div class="doc_text">
1037<p>The primitive types are the fundamental building blocks of the LLVM
1038system.</p>
1039
Chris Lattner86437612008-01-04 04:34:14 +0000[diff] [blame^]1040</div>
1041
Chris Lattner488772f2008-01-04 04:32:38 +0000[diff] [blame]1042<!-- _______________________________________________________________________ -->
1043<div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1044
1045<div class="doc_text">
1046 <table>
1047 <tbody>
1048 <tr><th>Type</th><th>Description</th></tr>
1049 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1050 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1051 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1052 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1053 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1054 </tbody>
1055 </table>
1056</div>
1057
1058<!-- _______________________________________________________________________ -->
1059<div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1060
1061<div class="doc_text">
1062<h5>Overview:</h5>
1063<p>The void type does not represent any value and has no size.</p>
1064
1065<h5>Syntax:</h5>
1066
1067<pre>
1068 void
1069</pre>
1070</div>
1071
1072<!-- _______________________________________________________________________ -->
1073<div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1074
1075<div class="doc_text">
1076<h5>Overview:</h5>
1077<p>The label type represents code labels.</p>
1078
1079<h5>Syntax:</h5>
1080
1081<pre>
1082 label
1083</pre>
1084</div>
1085
1086
1087<!-- ======================================================================= -->
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1088<div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
1089
1090<div class="doc_text">
1091
1092<p>The real power in LLVM comes from the derived types in the system.
1093This is what allows a programmer to represent arrays, functions,
1094pointers, and other useful types. Note that these derived types may be
1095recursive: For example, it is possible to have a two dimensional array.</p>
1096
1097</div>
1098
1099<!-- _______________________________________________________________________ -->
1100<div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1101
1102<div class="doc_text">
1103
1104<h5>Overview:</h5>
1105<p>The integer type is a very simple derived type that simply specifies an
1106arbitrary bit width for the integer type desired. Any bit width from 1 bit to
11072^23-1 (about 8 million) can be specified.</p>
1108
1109<h5>Syntax:</h5>
1110
1111<pre>
1112 iN
1113</pre>
1114
1115<p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1116value.</p>
1117
1118<h5>Examples:</h5>
1119<table class="layout">
Chris Lattner251ab812007-12-18 06:18:21 +0000[diff] [blame]1120 <tbody>
1121 <tr>
1122 <td><tt>i1</tt></td>
1123 <td>a single-bit integer.</td>
1124 </tr><tr>
1125 <td><tt>i32</tt></td>
1126 <td>a 32-bit integer.</td>
1127 </tr><tr>
1128 <td><tt>i1942652</tt></td>
1129 <td>a really big integer of over 1 million bits.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1130 </tr>
Chris Lattner251ab812007-12-18 06:18:21 +0000[diff] [blame]1131 </tbody>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1132</table>
1133</div>
1134
1135<!-- _______________________________________________________________________ -->
1136<div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
1137
1138<div class="doc_text">
1139
1140<h5>Overview:</h5>
1141
1142<p>The array type is a very simple derived type that arranges elements
1143sequentially in memory. The array type requires a size (number of
1144elements) and an underlying data type.</p>
1145
1146<h5>Syntax:</h5>
1147
1148<pre>
1149 [&lt;# elements&gt; x &lt;elementtype&gt;]
1150</pre>
1151
1152<p>The number of elements is a constant integer value; elementtype may
1153be any type with a size.</p>
1154
1155<h5>Examples:</h5>
1156<table class="layout">
1157 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1158 <td class="left"><tt>[40 x i32]</tt></td>
1159 <td class="left">Array of 40 32-bit integer values.</td>
1160 </tr>
1161 <tr class="layout">
1162 <td class="left"><tt>[41 x i32]</tt></td>
1163 <td class="left">Array of 41 32-bit integer values.</td>
1164 </tr>
1165 <tr class="layout">
1166 <td class="left"><tt>[4 x i8]</tt></td>
1167 <td class="left">Array of 4 8-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1168 </tr>
1169</table>
1170<p>Here are some examples of multidimensional arrays:</p>
1171<table class="layout">
1172 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1173 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1174 <td class="left">3x4 array of 32-bit integer values.</td>
1175 </tr>
1176 <tr class="layout">
1177 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1178 <td class="left">12x10 array of single precision floating point values.</td>
1179 </tr>
1180 <tr class="layout">
1181 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1182 <td class="left">2x3x4 array of 16-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1183 </tr>
1184</table>
1185
1186<p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1187length array. Normally, accesses past the end of an array are undefined in
1188LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1189As a special case, however, zero length arrays are recognized to be variable
1190length. This allows implementation of 'pascal style arrays' with the LLVM
1191type "{ i32, [0 x float]}", for example.</p>
1192
1193</div>
1194
1195<!-- _______________________________________________________________________ -->
1196<div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
1197<div class="doc_text">
1198<h5>Overview:</h5>
1199<p>The function type can be thought of as a function signature. It
1200consists of a return type and a list of formal parameter types.
1201Function types are usually used to build virtual function tables
1202(which are structures of pointers to functions), for indirect function
1203calls, and when defining a function.</p>
1204<p>
1205The return type of a function type cannot be an aggregate type.
1206</p>
1207<h5>Syntax:</h5>
1208<pre> &lt;returntype&gt; (&lt;parameter list&gt;)<br></pre>
1209<p>...where '<tt>&lt;parameter list&gt;</tt>' is a comma-separated list of type
1210specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1211which indicates that the function takes a variable number of arguments.
1212Variable argument functions can access their arguments with the <a
1213 href="#int_varargs">variable argument handling intrinsic</a> functions.</p>
1214<h5>Examples:</h5>
1215<table class="layout">
1216 <tr class="layout">
1217 <td class="left"><tt>i32 (i32)</tt></td>
1218 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
1219 </td>
1220 </tr><tr class="layout">
Reid Spencerf234bed2007-07-19 23:13:04 +0000[diff] [blame]1221 <td class="left"><tt>float&nbsp;(i16&nbsp;signext,&nbsp;i32&nbsp;*)&nbsp;*
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1222 </tt></td>
1223 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1224 an <tt>i16</tt> that should be sign extended and a
1225 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
1226 <tt>float</tt>.
1227 </td>
1228 </tr><tr class="layout">
1229 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1230 <td class="left">A vararg function that takes at least one
1231 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1232 which returns an integer. This is the signature for <tt>printf</tt> in
1233 LLVM.
1234 </td>
1235 </tr>
1236</table>
1237
1238</div>
1239<!-- _______________________________________________________________________ -->
1240<div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
1241<div class="doc_text">
1242<h5>Overview:</h5>
1243<p>The structure type is used to represent a collection of data members
1244together in memory. The packing of the field types is defined to match
1245the ABI of the underlying processor. The elements of a structure may
1246be any type that has a size.</p>
1247<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1248and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1249field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1250instruction.</p>
1251<h5>Syntax:</h5>
1252<pre> { &lt;type list&gt; }<br></pre>
1253<h5>Examples:</h5>
1254<table class="layout">
1255 <tr class="layout">
1256 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1257 <td class="left">A triple of three <tt>i32</tt> values</td>
1258 </tr><tr class="layout">
1259 <td class="left"><tt>{&nbsp;float,&nbsp;i32&nbsp;(i32)&nbsp;*&nbsp;}</tt></td>
1260 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1261 second element is a <a href="#t_pointer">pointer</a> to a
1262 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1263 an <tt>i32</tt>.</td>
1264 </tr>
1265</table>
1266</div>
1267
1268<!-- _______________________________________________________________________ -->
1269<div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1270</div>
1271<div class="doc_text">
1272<h5>Overview:</h5>
1273<p>The packed structure type is used to represent a collection of data members
1274together in memory. There is no padding between fields. Further, the alignment
1275of a packed structure is 1 byte. The elements of a packed structure may
1276be any type that has a size.</p>
1277<p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1278and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1279field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1280instruction.</p>
1281<h5>Syntax:</h5>
1282<pre> &lt; { &lt;type list&gt; } &gt; <br></pre>
1283<h5>Examples:</h5>
1284<table class="layout">
1285 <tr class="layout">
1286 <td class="left"><tt>&lt; { i32, i32, i32 } &gt;</tt></td>
1287 <td class="left">A triple of three <tt>i32</tt> values</td>
1288 </tr><tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1289 <td class="left"><tt>&lt; { float, i32 (i32)* } &gt;</tt></td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1290 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1291 second element is a <a href="#t_pointer">pointer</a> to a
1292 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1293 an <tt>i32</tt>.</td>
1294 </tr>
1295</table>
1296</div>
1297
1298<!-- _______________________________________________________________________ -->
1299<div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
1300<div class="doc_text">
1301<h5>Overview:</h5>
1302<p>As in many languages, the pointer type represents a pointer or
Christopher Lambdd0049d2007-12-11 09:31:00 +0000[diff] [blame]1303reference to another object, which must live in memory. Pointer types may have
1304an optional address space attribute defining the target-specific numbered
1305address space where the pointed-to object resides. The default address space is
1306zero.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1307<h5>Syntax:</h5>
1308<pre> &lt;type&gt; *<br></pre>
1309<h5>Examples:</h5>
1310<table class="layout">
1311 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1312 <td class="left"><tt>[4x i32]*</tt></td>
1313 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1314 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1315 </tr>
1316 <tr class="layout">
1317 <td class="left"><tt>i32 (i32 *) *</tt></td>
1318 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1319 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1320 <tt>i32</tt>.</td>
1321 </tr>
1322 <tr class="layout">
1323 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1324 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1325 that resides in address space #5.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1326 </tr>
1327</table>
1328</div>
1329
1330<!-- _______________________________________________________________________ -->
1331<div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
1332<div class="doc_text">
1333
1334<h5>Overview:</h5>
1335
1336<p>A vector type is a simple derived type that represents a vector
1337of elements. Vector types are used when multiple primitive data
1338are operated in parallel using a single instruction (SIMD).
1339A vector type requires a size (number of
1340elements) and an underlying primitive data type. Vectors must have a power
1341of two length (1, 2, 4, 8, 16 ...). Vector types are
1342considered <a href="#t_firstclass">first class</a>.</p>
1343
1344<h5>Syntax:</h5>
1345
1346<pre>
1347 &lt; &lt;# elements&gt; x &lt;elementtype&gt; &gt;
1348</pre>
1349
1350<p>The number of elements is a constant integer value; elementtype may
1351be any integer or floating point type.</p>
1352
1353<h5>Examples:</h5>
1354
1355<table class="layout">
1356 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1357 <td class="left"><tt>&lt;4 x i32&gt;</tt></td>
1358 <td class="left">Vector of 4 32-bit integer values.</td>
1359 </tr>
1360 <tr class="layout">
1361 <td class="left"><tt>&lt;8 x float&gt;</tt></td>
1362 <td class="left">Vector of 8 32-bit floating-point values.</td>
1363 </tr>
1364 <tr class="layout">
1365 <td class="left"><tt>&lt;2 x i64&gt;</tt></td>
1366 <td class="left">Vector of 2 64-bit integer values.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1367 </tr>
1368</table>
1369</div>
1370
1371<!-- _______________________________________________________________________ -->
1372<div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1373<div class="doc_text">
1374
1375<h5>Overview:</h5>
1376
1377<p>Opaque types are used to represent unknown types in the system. This
Gordon Henriksenda0706e2007-10-14 00:34:53 +0000[diff] [blame]1378corresponds (for example) to the C notion of a forward declared structure type.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1379In LLVM, opaque types can eventually be resolved to any type (not just a
1380structure type).</p>
1381
1382<h5>Syntax:</h5>
1383
1384<pre>
1385 opaque
1386</pre>
1387
1388<h5>Examples:</h5>
1389
1390<table class="layout">
1391 <tr class="layout">
Chris Lattner7311d222007-12-19 05:04:11 +0000[diff] [blame]1392 <td class="left"><tt>opaque</tt></td>
1393 <td class="left">An opaque type.</td>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1394 </tr>
1395</table>
1396</div>
1397
1398
1399<!-- *********************************************************************** -->
1400<div class="doc_section"> <a name="constants">Constants</a> </div>
1401<!-- *********************************************************************** -->
1402
1403<div class="doc_text">
1404
1405<p>LLVM has several different basic types of constants. This section describes
1406them all and their syntax.</p>
1407
1408</div>
1409
1410<!-- ======================================================================= -->
1411<div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
1412
1413<div class="doc_text">
1414
1415<dl>
1416 <dt><b>Boolean constants</b></dt>
1417
1418 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
1419 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
1420 </dd>
1421
1422 <dt><b>Integer constants</b></dt>
1423
1424 <dd>Standard integers (such as '4') are constants of the <a
1425 href="#t_integer">integer</a> type. Negative numbers may be used with
1426 integer types.
1427 </dd>
1428
1429 <dt><b>Floating point constants</b></dt>
1430
1431 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1432 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
1433 notation (see below). Floating point constants must have a <a
1434 href="#t_floating">floating point</a> type. </dd>
1435
1436 <dt><b>Null pointer constants</b></dt>
1437
1438 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
1439 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1440
1441</dl>
1442
1443<p>The one non-intuitive notation for constants is the optional hexadecimal form
1444of floating point constants. For example, the form '<tt>double
14450x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
14464.5e+15</tt>'. The only time hexadecimal floating point constants are required
1447(and the only time that they are generated by the disassembler) is when a
1448floating point constant must be emitted but it cannot be represented as a
1449decimal floating point number. For example, NaN's, infinities, and other
1450special values are represented in their IEEE hexadecimal format so that
1451assembly and disassembly do not cause any bits to change in the constants.</p>
1452
1453</div>
1454
1455<!-- ======================================================================= -->
1456<div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1457</div>
1458
1459<div class="doc_text">
1460<p>Aggregate constants arise from aggregation of simple constants
1461and smaller aggregate constants.</p>
1462
1463<dl>
1464 <dt><b>Structure constants</b></dt>
1465
1466 <dd>Structure constants are represented with notation similar to structure
1467 type definitions (a comma separated list of elements, surrounded by braces
Chris Lattner6c8de962007-12-25 20:34:52 +0000[diff] [blame]1468 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
1469 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]1470 must have <a href="#t_struct">structure type</a>, and the number and
1471 types of elements must match those specified by the type.
1472 </dd>
1473
1474 <dt><b>Array constants</b></dt>
1475
1476 <dd>Array constants are represented with notation similar to array type
1477 definitions (a comma separated list of elements, surrounded by square brackets
1478 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
1479 constants must have <a href="#t_array">array type</a>, and the number and
1480 types of elements must match those specified by the type.
1481 </dd>
1482
1483 <dt><b>Vector constants</b></dt>
1484
1485 <dd>Vector constants are represented with notation similar to vector type
1486 definitions (a comma separated list of elements, surrounded by
1487 less-than/greater-than's (<tt>&lt;&gt;</tt>)). For example: "<tt>&lt; i32 42,
1488 i32 11, i32 74, i32 100 &gt;</tt>". Vector constants must have <a
1489 href="#t_vector">vector type</a>, and the number and types of elements must
1490 match those specified by the type.
1491 </dd>
1492
1493 <dt><b>Zero initialization</b></dt>
1494
1495 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1496 value to zero of <em>any</em> type, including scalar and aggregate types.
1497 This is often used to avoid having to print large zero initializers (e.g. for
1498 large arrays) and is always exactly equivalent to using explicit zero
1499 initializers.
1500 </dd>
1501</dl>
1502
1503</div>
1504
1505<!-- ======================================================================= -->
1506<div class="doc_subsection">
1507 <a name="globalconstants">Global Variable and Function Addresses</a>
1508</div>
1509
1510<div class="doc_text">
1511
1512<p>The addresses of <a href="#globalvars">global variables</a> and <a
1513href="#functionstructure">functions</a> are always implicitly valid (link-time)
1514constants. These constants are explicitly referenced when the <a
1515href="#identifiers">identifier for the global</a> is used and always have <a
1516href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1517file:</p>
1518
1519<div class="doc_code">
1520<pre>
1521@X = global i32 17
1522@Y = global i32 42
1523@Z = global [2 x i32*] [ i32* @X, i32* @Y ]
1524</pre>
1525</div>
1526
1527</div>
1528
1529<!-- ======================================================================= -->
1530<div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
1531<div class="doc_text">
1532 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
1533 no specific value. Undefined values may be of any type and be used anywhere
1534 a constant is permitted.</p>
1535
1536 <p>Undefined values indicate to the compiler that the program is well defined
1537 no matter what value is used, giving the compiler more freedom to optimize.
1538 </p>
1539</div>
1540
1541<!-- ======================================================================= -->
1542<div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1543</div>
1544
1545<div class="doc_text">
1546
1547<p>Constant expressions are used to allow expressions involving other constants
1548to be used as constants. Constant expressions may be of any <a
1549href="#t_firstclass">first class</a> type and may involve any LLVM operation
1550that does not have side effects (e.g. load and call are not supported). The
1551following is the syntax for constant expressions:</p>
1552
1553<dl>
1554 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1555 <dd>Truncate a constant to another type. The bit size of CST must be larger
1556 than the bit size of TYPE. Both types must be integers.</dd>
1557
1558 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1559 <dd>Zero extend a constant to another type. The bit size of CST must be
1560 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1561
1562 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1563 <dd>Sign extend a constant to another type. The bit size of CST must be
1564 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1565
1566 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1567 <dd>Truncate a floating point constant to another floating point type. The
1568 size of CST must be larger than the size of TYPE. Both types must be
1569 floating point.</dd>
1570
1571 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1572 <dd>Floating point extend a constant to another type. The size of CST must be
1573 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1574
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]1575 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1576 <dd>Convert a floating point constant to the corresponding unsigned integer
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1577 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1578 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1579 of the same number of elements. If the value won't fit in the integer type,
1580 the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1581
1582 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
1583 <dd>Convert a floating point constant to the corresponding signed integer
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1584 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1585 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1586 of the same number of elements. If the value won't fit in the integer type,
1587 the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1588
1589 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
1590 <dd>Convert an unsigned integer constant to the corresponding floating point
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1591 constant. TYPE must be a scalar or vector floating point type. CST must be of
1592 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1593 of the same number of elements. If the value won't fit in the floating point
1594 type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1595
1596 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
1597 <dd>Convert a signed integer constant to the corresponding floating point
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]1598 constant. TYPE must be a scalar or vector floating point type. CST must be of
1599 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1600 of the same number of elements. If the value won't fit in the floating point
1601 type, the results are undefined.</dd>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]1602
1603 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1604 <dd>Convert a pointer typed constant to the corresponding integer constant
1605 TYPE must be an integer type. CST must be of pointer type. The CST value is
1606 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1607
1608 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1609 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1610 pointer type. CST must be of integer type. The CST value is zero extended,
1611 truncated, or unchanged to make it fit in a pointer size. This one is
1612 <i>really</i> dangerous!</dd>
1613
1614 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
1615 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1616 identical (same number of bits). The conversion is done as if the CST value
1617 was stored to memory and read back as TYPE. In other words, no bits change
1618 with this operator, just the type. This can be used for conversion of
1619 vector types to any other type, as long as they have the same bit width. For
1620 pointers it is only valid to cast to another pointer type.
1621 </dd>
1622
1623 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1624
1625 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1626 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1627 instruction, the index list may have zero or more indexes, which are required
1628 to make sense for the type of "CSTPTR".</dd>
1629
1630 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1631
1632 <dd>Perform the <a href="#i_select">select operation</a> on
1633 constants.</dd>
1634
1635 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1636 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1637
1638 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1639 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
1640
1641 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1642
1643 <dd>Perform the <a href="#i_extractelement">extractelement
1644 operation</a> on constants.
1645
1646 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1647
1648 <dd>Perform the <a href="#i_insertelement">insertelement
1649 operation</a> on constants.</dd>
1650
1651
1652 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1653
1654 <dd>Perform the <a href="#i_shufflevector">shufflevector
1655 operation</a> on constants.</dd>
1656
1657 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1658
1659 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1660 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
1661 binary</a> operations. The constraints on operands are the same as those for
1662 the corresponding instruction (e.g. no bitwise operations on floating point
1663 values are allowed).</dd>
1664</dl>
1665</div>
1666
1667<!-- *********************************************************************** -->
1668<div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1669<!-- *********************************************************************** -->
1670
1671<!-- ======================================================================= -->
1672<div class="doc_subsection">
1673<a name="inlineasm">Inline Assembler Expressions</a>
1674</div>
1675
1676<div class="doc_text">
1677
1678<p>
1679LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
1680Module-Level Inline Assembly</a>) through the use of a special value. This
1681value represents the inline assembler as a string (containing the instructions
1682to emit), a list of operand constraints (stored as a string), and a flag that
1683indicates whether or not the inline asm expression has side effects. An example
1684inline assembler expression is:
1685</p>
1686
1687<div class="doc_code">
1688<pre>
1689i32 (i32) asm "bswap $0", "=r,r"
1690</pre>
1691</div>
1692
1693<p>
1694Inline assembler expressions may <b>only</b> be used as the callee operand of
1695a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
1696</p>
1697
1698<div class="doc_code">
1699<pre>
1700%X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
1701</pre>
1702</div>
1703
1704<p>
1705Inline asms with side effects not visible in the constraint list must be marked
1706as having side effects. This is done through the use of the
1707'<tt>sideeffect</tt>' keyword, like so:
1708</p>
1709
1710<div class="doc_code">
1711<pre>
1712call void asm sideeffect "eieio", ""()
1713</pre>
1714</div>
1715
1716<p>TODO: The format of the asm and constraints string still need to be
1717documented here. Constraints on what can be done (e.g. duplication, moving, etc
1718need to be documented).
1719</p>
1720
1721</div>
1722
1723<!-- *********************************************************************** -->
1724<div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
1725<!-- *********************************************************************** -->
1726
1727<div class="doc_text">
1728
1729<p>The LLVM instruction set consists of several different
1730classifications of instructions: <a href="#terminators">terminator
1731instructions</a>, <a href="#binaryops">binary instructions</a>,
1732<a href="#bitwiseops">bitwise binary instructions</a>, <a
1733 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
1734instructions</a>.</p>
1735
1736</div>
1737
1738<!-- ======================================================================= -->
1739<div class="doc_subsection"> <a name="terminators">Terminator
1740Instructions</a> </div>
1741
1742<div class="doc_text">
1743
1744<p>As mentioned <a href="#functionstructure">previously</a>, every
1745basic block in a program ends with a "Terminator" instruction, which
1746indicates which block should be executed after the current block is
1747finished. These terminator instructions typically yield a '<tt>void</tt>'
1748value: they produce control flow, not values (the one exception being
1749the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
1750<p>There are six different terminator instructions: the '<a
1751 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
1752instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
1753the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
1754 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
1755 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
1756
1757</div>
1758
1759<!-- _______________________________________________________________________ -->
1760<div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
1761Instruction</a> </div>
1762<div class="doc_text">
1763<h5>Syntax:</h5>
1764<pre> ret &lt;type&gt; &lt;value&gt; <i>; Return a value from a non-void function</i>
1765 ret void <i>; Return from void function</i>
1766</pre>
1767<h5>Overview:</h5>
1768<p>The '<tt>ret</tt>' instruction is used to return control flow (and a
1769value) from a function back to the caller.</p>
1770<p>There are two forms of the '<tt>ret</tt>' instruction: one that
1771returns a value and then causes control flow, and one that just causes
1772control flow to occur.</p>
1773<h5>Arguments:</h5>
1774<p>The '<tt>ret</tt>' instruction may return any '<a
1775 href="#t_firstclass">first class</a>' type. Notice that a function is
1776not <a href="#wellformed">well formed</a> if there exists a '<tt>ret</tt>'
1777instruction inside of the function that returns a value that does not
1778match the return type of the function.</p>
1779<h5>Semantics:</h5>
1780<p>When the '<tt>ret</tt>' instruction is executed, control flow
1781returns back to the calling function's context. If the caller is a "<a
1782 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
1783the instruction after the call. If the caller was an "<a
1784 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
1785at the beginning of the "normal" destination block. If the instruction
1786returns a value, that value shall set the call or invoke instruction's
1787return value.</p>
1788<h5>Example:</h5>
1789<pre> ret i32 5 <i>; Return an integer value of 5</i>
1790 ret void <i>; Return from a void function</i>
1791</pre>
1792</div>
1793<!-- _______________________________________________________________________ -->
1794<div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
1795<div class="doc_text">
1796<h5>Syntax:</h5>
1797<pre> br i1 &lt;cond&gt;, label &lt;iftrue&gt;, label &lt;iffalse&gt;<br> br label &lt;dest&gt; <i>; Unconditional branch</i>
1798</pre>
1799<h5>Overview:</h5>
1800<p>The '<tt>br</tt>' instruction is used to cause control flow to
1801transfer to a different basic block in the current function. There are
1802two forms of this instruction, corresponding to a conditional branch
1803and an unconditional branch.</p>
1804<h5>Arguments:</h5>
1805<p>The conditional branch form of the '<tt>br</tt>' instruction takes a
1806single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
1807unconditional form of the '<tt>br</tt>' instruction takes a single
1808'<tt>label</tt>' value as a target.</p>
1809<h5>Semantics:</h5>
1810<p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
1811argument is evaluated. If the value is <tt>true</tt>, control flows
1812to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
1813control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
1814<h5>Example:</h5>
1815<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
1816 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
1817</div>
1818<!-- _______________________________________________________________________ -->
1819<div class="doc_subsubsection">
1820 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
1821</div>
1822
1823<div class="doc_text">
1824<h5>Syntax:</h5>
1825
1826<pre>
1827 switch &lt;intty&gt; &lt;value&gt;, label &lt;defaultdest&gt; [ &lt;intty&gt; &lt;val&gt;, label &lt;dest&gt; ... ]
1828</pre>
1829
1830<h5>Overview:</h5>
1831
1832<p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
1833several different places. It is a generalization of the '<tt>br</tt>'
1834instruction, allowing a branch to occur to one of many possible
1835destinations.</p>
1836
1837
1838<h5>Arguments:</h5>
1839
1840<p>The '<tt>switch</tt>' instruction uses three parameters: an integer
1841comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
1842an array of pairs of comparison value constants and '<tt>label</tt>'s. The
1843table is not allowed to contain duplicate constant entries.</p>
1844
1845<h5>Semantics:</h5>
1846
1847<p>The <tt>switch</tt> instruction specifies a table of values and
1848destinations. When the '<tt>switch</tt>' instruction is executed, this
1849table is searched for the given value. If the value is found, control flow is
1850transfered to the corresponding destination; otherwise, control flow is
1851transfered to the default destination.</p>
1852
1853<h5>Implementation:</h5>
1854
1855<p>Depending on properties of the target machine and the particular
1856<tt>switch</tt> instruction, this instruction may be code generated in different
1857ways. For example, it could be generated as a series of chained conditional
1858branches or with a lookup table.</p>
1859
1860<h5>Example:</h5>
1861
1862<pre>
1863 <i>; Emulate a conditional br instruction</i>
1864 %Val = <a href="#i_zext">zext</a> i1 %value to i32
1865 switch i32 %Val, label %truedest [i32 0, label %falsedest ]
1866
1867 <i>; Emulate an unconditional br instruction</i>
1868 switch i32 0, label %dest [ ]
1869
1870 <i>; Implement a jump table:</i>
1871 switch i32 %val, label %otherwise [ i32 0, label %onzero
1872 i32 1, label %onone
1873 i32 2, label %ontwo ]
1874</pre>
1875</div>
1876
1877<!-- _______________________________________________________________________ -->
1878<div class="doc_subsubsection">
1879 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
1880</div>
1881
1882<div class="doc_text">
1883
1884<h5>Syntax:</h5>
1885
1886<pre>
1887 &lt;result&gt; = invoke [<a href="#callingconv">cconv</a>] &lt;ptr to function ty&gt; %&lt;function ptr val&gt;(&lt;function args&gt;)
1888 to label &lt;normal label&gt; unwind label &lt;exception label&gt;
1889</pre>
1890
1891<h5>Overview:</h5>
1892
1893<p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
1894function, with the possibility of control flow transfer to either the
1895'<tt>normal</tt>' label or the
1896'<tt>exception</tt>' label. If the callee function returns with the
1897"<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
1898"normal" label. If the callee (or any indirect callees) returns with the "<a
1899href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
1900continued at the dynamically nearest "exception" label.</p>
1901
1902<h5>Arguments:</h5>
1903
1904<p>This instruction requires several arguments:</p>
1905
1906<ol>
1907 <li>
1908 The optional "cconv" marker indicates which <a href="#callingconv">calling
1909 convention</a> the call should use. If none is specified, the call defaults
1910 to using C calling conventions.
1911 </li>
1912 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
1913 function value being invoked. In most cases, this is a direct function
1914 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
1915 an arbitrary pointer to function value.
1916 </li>
1917
1918 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
1919 function to be invoked. </li>
1920
1921 <li>'<tt>function args</tt>': argument list whose types match the function
1922 signature argument types. If the function signature indicates the function
1923 accepts a variable number of arguments, the extra arguments can be
1924 specified. </li>
1925
1926 <li>'<tt>normal label</tt>': the label reached when the called function
1927 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
1928
1929 <li>'<tt>exception label</tt>': the label reached when a callee returns with
1930 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
1931
1932</ol>
1933
1934<h5>Semantics:</h5>
1935
1936<p>This instruction is designed to operate as a standard '<tt><a
1937href="#i_call">call</a></tt>' instruction in most regards. The primary
1938difference is that it establishes an association with a label, which is used by
1939the runtime library to unwind the stack.</p>
1940
1941<p>This instruction is used in languages with destructors to ensure that proper
1942cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
1943exception. Additionally, this is important for implementation of
1944'<tt>catch</tt>' clauses in high-level languages that support them.</p>
1945
1946<h5>Example:</h5>
1947<pre>
1948 %retval = invoke i32 %Test(i32 15) to label %Continue
1949 unwind label %TestCleanup <i>; {i32}:retval set</i>
1950 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Test(i32 15) to label %Continue
1951 unwind label %TestCleanup <i>; {i32}:retval set</i>
1952</pre>
1953</div>
1954
1955
1956<!-- _______________________________________________________________________ -->
1957
1958<div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
1959Instruction</a> </div>
1960
1961<div class="doc_text">
1962
1963<h5>Syntax:</h5>
1964<pre>
1965 unwind
1966</pre>
1967
1968<h5>Overview:</h5>
1969
1970<p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
1971at the first callee in the dynamic call stack which used an <a
1972href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
1973primarily used to implement exception handling.</p>
1974
1975<h5>Semantics:</h5>
1976
1977<p>The '<tt>unwind</tt>' intrinsic causes execution of the current function to
1978immediately halt. The dynamic call stack is then searched for the first <a
1979href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
1980execution continues at the "exceptional" destination block specified by the
1981<tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
1982dynamic call chain, undefined behavior results.</p>
1983</div>
1984
1985<!-- _______________________________________________________________________ -->
1986
1987<div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
1988Instruction</a> </div>
1989
1990<div class="doc_text">
1991
1992<h5>Syntax:</h5>
1993<pre>
1994 unreachable
1995</pre>
1996
1997<h5>Overview:</h5>
1998
1999<p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
2000instruction is used to inform the optimizer that a particular portion of the
2001code is not reachable. This can be used to indicate that the code after a
2002no-return function cannot be reached, and other facts.</p>
2003
2004<h5>Semantics:</h5>
2005
2006<p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
2007</div>
2008
2009
2010
2011<!-- ======================================================================= -->
2012<div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
2013<div class="doc_text">
2014<p>Binary operators are used to do most of the computation in a
2015program. They require two operands, execute an operation on them, and
2016produce a single value. The operands might represent
2017multiple data, as is the case with the <a href="#t_vector">vector</a> data type.
2018The result value of a binary operator is not
2019necessarily the same type as its operands.</p>
2020<p>There are several different binary operators:</p>
2021</div>
2022<!-- _______________________________________________________________________ -->
2023<div class="doc_subsubsection"> <a name="i_add">'<tt>add</tt>'
2024Instruction</a> </div>
2025<div class="doc_text">
2026<h5>Syntax:</h5>
2027<pre> &lt;result&gt; = add &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2028</pre>
2029<h5>Overview:</h5>
2030<p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
2031<h5>Arguments:</h5>
2032<p>The two arguments to the '<tt>add</tt>' instruction must be either <a
2033 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a> values.
2034 This instruction can also take <a href="#t_vector">vector</a> versions of the values.
2035Both arguments must have identical types.</p>
2036<h5>Semantics:</h5>
2037<p>The value produced is the integer or floating point sum of the two
2038operands.</p>
2039<h5>Example:</h5>
2040<pre> &lt;result&gt; = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
2041</pre>
2042</div>
2043<!-- _______________________________________________________________________ -->
2044<div class="doc_subsubsection"> <a name="i_sub">'<tt>sub</tt>'
2045Instruction</a> </div>
2046<div class="doc_text">
2047<h5>Syntax:</h5>
2048<pre> &lt;result&gt; = sub &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2049</pre>
2050<h5>Overview:</h5>
2051<p>The '<tt>sub</tt>' instruction returns the difference of its two
2052operands.</p>
2053<p>Note that the '<tt>sub</tt>' instruction is used to represent the '<tt>neg</tt>'
2054instruction present in most other intermediate representations.</p>
2055<h5>Arguments:</h5>
2056<p>The two arguments to the '<tt>sub</tt>' instruction must be either <a
2057 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
2058values.
2059This instruction can also take <a href="#t_vector">vector</a> versions of the values.
2060Both arguments must have identical types.</p>
2061<h5>Semantics:</h5>
2062<p>The value produced is the integer or floating point difference of
2063the two operands.</p>
2064<h5>Example:</h5>
2065<pre>
2066 &lt;result&gt; = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
2067 &lt;result&gt; = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
2068</pre>
2069</div>
2070<!-- _______________________________________________________________________ -->
2071<div class="doc_subsubsection"> <a name="i_mul">'<tt>mul</tt>'
2072Instruction</a> </div>
2073<div class="doc_text">
2074<h5>Syntax:</h5>
2075<pre> &lt;result&gt; = mul &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2076</pre>
2077<h5>Overview:</h5>
2078<p>The '<tt>mul</tt>' instruction returns the product of its two
2079operands.</p>
2080<h5>Arguments:</h5>
2081<p>The two arguments to the '<tt>mul</tt>' instruction must be either <a
2082 href="#t_integer">integer</a> or <a href="#t_floating">floating point</a>
2083values.
2084This instruction can also take <a href="#t_vector">vector</a> versions of the values.
2085Both arguments must have identical types.</p>
2086<h5>Semantics:</h5>
2087<p>The value produced is the integer or floating point product of the
2088two operands.</p>
2089<p>Because the operands are the same width, the result of an integer
2090multiplication is the same whether the operands should be deemed unsigned or
2091signed.</p>
2092<h5>Example:</h5>
2093<pre> &lt;result&gt; = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
2094</pre>
2095</div>
2096<!-- _______________________________________________________________________ -->
2097<div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2098</a></div>
2099<div class="doc_text">
2100<h5>Syntax:</h5>
2101<pre> &lt;result&gt; = udiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2102</pre>
2103<h5>Overview:</h5>
2104<p>The '<tt>udiv</tt>' instruction returns the quotient of its two
2105operands.</p>
2106<h5>Arguments:</h5>
2107<p>The two arguments to the '<tt>udiv</tt>' instruction must be
2108<a href="#t_integer">integer</a> values. Both arguments must have identical
2109types. This instruction can also take <a href="#t_vector">vector</a> versions
2110of the values in which case the elements must be integers.</p>
2111<h5>Semantics:</h5>
2112<p>The value produced is the unsigned integer quotient of the two operands. This
2113instruction always performs an unsigned division operation, regardless of
2114whether the arguments are unsigned or not.</p>
2115<h5>Example:</h5>
2116<pre> &lt;result&gt; = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
2117</pre>
2118</div>
2119<!-- _______________________________________________________________________ -->
2120<div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2121</a> </div>
2122<div class="doc_text">
2123<h5>Syntax:</h5>
2124<pre> &lt;result&gt; = sdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2125</pre>
2126<h5>Overview:</h5>
2127<p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
2128operands.</p>
2129<h5>Arguments:</h5>
2130<p>The two arguments to the '<tt>sdiv</tt>' instruction must be
2131<a href="#t_integer">integer</a> values. Both arguments must have identical
2132types. This instruction can also take <a href="#t_vector">vector</a> versions
2133of the values in which case the elements must be integers.</p>
2134<h5>Semantics:</h5>
2135<p>The value produced is the signed integer quotient of the two operands. This
2136instruction always performs a signed division operation, regardless of whether
2137the arguments are signed or not.</p>
2138<h5>Example:</h5>
2139<pre> &lt;result&gt; = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
2140</pre>
2141</div>
2142<!-- _______________________________________________________________________ -->
2143<div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
2144Instruction</a> </div>
2145<div class="doc_text">
2146<h5>Syntax:</h5>
2147<pre> &lt;result&gt; = fdiv &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2148</pre>
2149<h5>Overview:</h5>
2150<p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
2151operands.</p>
2152<h5>Arguments:</h5>
2153<p>The two arguments to the '<tt>fdiv</tt>' instruction must be
2154<a href="#t_floating">floating point</a> values. Both arguments must have
2155identical types. This instruction can also take <a href="#t_vector">vector</a>
2156versions of floating point values.</p>
2157<h5>Semantics:</h5>
2158<p>The value produced is the floating point quotient of the two operands.</p>
2159<h5>Example:</h5>
2160<pre> &lt;result&gt; = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
2161</pre>
2162</div>
2163<!-- _______________________________________________________________________ -->
2164<div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2165</div>
2166<div class="doc_text">
2167<h5>Syntax:</h5>
2168<pre> &lt;result&gt; = urem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2169</pre>
2170<h5>Overview:</h5>
2171<p>The '<tt>urem</tt>' instruction returns the remainder from the
2172unsigned division of its two arguments.</p>
2173<h5>Arguments:</h5>
2174<p>The two arguments to the '<tt>urem</tt>' instruction must be
2175<a href="#t_integer">integer</a> values. Both arguments must have identical
Dan Gohman3e3fd8c2007-11-05 23:35:22 +0000[diff] [blame]2176types. This instruction can also take <a href="#t_vector">vector</a> versions
2177of the values in which case the elements must be integers.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2178<h5>Semantics:</h5>
2179<p>This instruction returns the unsigned integer <i>remainder</i> of a division.
2180This instruction always performs an unsigned division to get the remainder,
2181regardless of whether the arguments are unsigned or not.</p>
2182<h5>Example:</h5>
2183<pre> &lt;result&gt; = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2184</pre>
2185
2186</div>
2187<!-- _______________________________________________________________________ -->
2188<div class="doc_subsubsection"> <a name="i_srem">'<tt>srem</tt>'
2189Instruction</a> </div>
2190<div class="doc_text">
2191<h5>Syntax:</h5>
2192<pre> &lt;result&gt; = srem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2193</pre>
2194<h5>Overview:</h5>
2195<p>The '<tt>srem</tt>' instruction returns the remainder from the
Dan Gohman3e3fd8c2007-11-05 23:35:22 +0000[diff] [blame]2196signed division of its two operands. This instruction can also take
2197<a href="#t_vector">vector</a> versions of the values in which case
2198the elements must be integers.</p>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]2199
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2200<h5>Arguments:</h5>
2201<p>The two arguments to the '<tt>srem</tt>' instruction must be
2202<a href="#t_integer">integer</a> values. Both arguments must have identical
2203types.</p>
2204<h5>Semantics:</h5>
2205<p>This instruction returns the <i>remainder</i> of a division (where the result
2206has the same sign as the dividend, <tt>var1</tt>), not the <i>modulo</i>
2207operator (where the result has the same sign as the divisor, <tt>var2</tt>) of
2208a value. For more information about the difference, see <a
2209 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
2210Math Forum</a>. For a table of how this is implemented in various languages,
2211please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
2212Wikipedia: modulo operation</a>.</p>
2213<h5>Example:</h5>
2214<pre> &lt;result&gt; = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2215</pre>
2216
2217</div>
2218<!-- _______________________________________________________________________ -->
2219<div class="doc_subsubsection"> <a name="i_frem">'<tt>frem</tt>'
2220Instruction</a> </div>
2221<div class="doc_text">
2222<h5>Syntax:</h5>
2223<pre> &lt;result&gt; = frem &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2224</pre>
2225<h5>Overview:</h5>
2226<p>The '<tt>frem</tt>' instruction returns the remainder from the
2227division of its two operands.</p>
2228<h5>Arguments:</h5>
2229<p>The two arguments to the '<tt>frem</tt>' instruction must be
2230<a href="#t_floating">floating point</a> values. Both arguments must have
Dan Gohman3e3fd8c2007-11-05 23:35:22 +0000[diff] [blame]2231identical types. This instruction can also take <a href="#t_vector">vector</a>
2232versions of floating point values.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2233<h5>Semantics:</h5>
2234<p>This instruction returns the <i>remainder</i> of a division.</p>
2235<h5>Example:</h5>
2236<pre> &lt;result&gt; = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
2237</pre>
2238</div>
2239
2240<!-- ======================================================================= -->
2241<div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2242Operations</a> </div>
2243<div class="doc_text">
2244<p>Bitwise binary operators are used to do various forms of
2245bit-twiddling in a program. They are generally very efficient
2246instructions and can commonly be strength reduced from other
2247instructions. They require two operands, execute an operation on them,
2248and produce a single value. The resulting value of the bitwise binary
2249operators is always the same type as its first operand.</p>
2250</div>
2251
2252<!-- _______________________________________________________________________ -->
2253<div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2254Instruction</a> </div>
2255<div class="doc_text">
2256<h5>Syntax:</h5>
2257<pre> &lt;result&gt; = shl &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2258</pre>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2259
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2260<h5>Overview:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2261
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2262<p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2263the left a specified number of bits.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2264
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2265<h5>Arguments:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2266
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2267<p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
2268 href="#t_integer">integer</a> type.</p>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2269
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2270<h5>Semantics:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2271
2272<p>The value produced is <tt>var1</tt> * 2<sup><tt>var2</tt></sup>. If
2273<tt>var2</tt> is (statically or dynamically) equal to or larger than the number
2274of bits in <tt>var1</tt>, the result is undefined.</p>
2275
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2276<h5>Example:</h5><pre>
2277 &lt;result&gt; = shl i32 4, %var <i>; yields {i32}: 4 &lt;&lt; %var</i>
2278 &lt;result&gt; = shl i32 4, 2 <i>; yields {i32}: 16</i>
2279 &lt;result&gt; = shl i32 1, 10 <i>; yields {i32}: 1024</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2280 &lt;result&gt; = shl i32 1, 32 <i>; undefined</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2281</pre>
2282</div>
2283<!-- _______________________________________________________________________ -->
2284<div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2285Instruction</a> </div>
2286<div class="doc_text">
2287<h5>Syntax:</h5>
2288<pre> &lt;result&gt; = lshr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2289</pre>
2290
2291<h5>Overview:</h5>
2292<p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2293operand shifted to the right a specified number of bits with zero fill.</p>
2294
2295<h5>Arguments:</h5>
2296<p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
2297<a href="#t_integer">integer</a> type.</p>
2298
2299<h5>Semantics:</h5>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2300
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2301<p>This instruction always performs a logical shift right operation. The most
2302significant bits of the result will be filled with zero bits after the
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2303shift. If <tt>var2</tt> is (statically or dynamically) equal to or larger than
2304the number of bits in <tt>var1</tt>, the result is undefined.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2305
2306<h5>Example:</h5>
2307<pre>
2308 &lt;result&gt; = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2309 &lt;result&gt; = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2310 &lt;result&gt; = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2311 &lt;result&gt; = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2312 &lt;result&gt; = lshr i32 1, 32 <i>; undefined</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2313</pre>
2314</div>
2315
2316<!-- _______________________________________________________________________ -->
2317<div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2318Instruction</a> </div>
2319<div class="doc_text">
2320
2321<h5>Syntax:</h5>
2322<pre> &lt;result&gt; = ashr &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2323</pre>
2324
2325<h5>Overview:</h5>
2326<p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2327operand shifted to the right a specified number of bits with sign extension.</p>
2328
2329<h5>Arguments:</h5>
2330<p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
2331<a href="#t_integer">integer</a> type.</p>
2332
2333<h5>Semantics:</h5>
2334<p>This instruction always performs an arithmetic shift right operation,
2335The most significant bits of the result will be filled with the sign bit
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2336of <tt>var1</tt>. If <tt>var2</tt> is (statically or dynamically) equal to or
2337larger than the number of bits in <tt>var1</tt>, the result is undefined.
2338</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2339
2340<h5>Example:</h5>
2341<pre>
2342 &lt;result&gt; = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2343 &lt;result&gt; = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2344 &lt;result&gt; = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2345 &lt;result&gt; = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
Chris Lattnerd939d9f2007-10-03 21:01:14 +0000[diff] [blame]2346 &lt;result&gt; = ashr i32 1, 32 <i>; undefined</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2347</pre>
2348</div>
2349
2350<!-- _______________________________________________________________________ -->
2351<div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2352Instruction</a> </div>
2353<div class="doc_text">
2354<h5>Syntax:</h5>
2355<pre> &lt;result&gt; = and &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2356</pre>
2357<h5>Overview:</h5>
2358<p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2359its two operands.</p>
2360<h5>Arguments:</h5>
2361<p>The two arguments to the '<tt>and</tt>' instruction must be <a
2362 href="#t_integer">integer</a> values. Both arguments must have
2363identical types.</p>
2364<h5>Semantics:</h5>
2365<p>The truth table used for the '<tt>and</tt>' instruction is:</p>
2366<p> </p>
2367<div style="align: center">
2368<table border="1" cellspacing="0" cellpadding="4">
2369 <tbody>
2370 <tr>
2371 <td>In0</td>
2372 <td>In1</td>
2373 <td>Out</td>
2374 </tr>
2375 <tr>
2376 <td>0</td>
2377 <td>0</td>
2378 <td>0</td>
2379 </tr>
2380 <tr>
2381 <td>0</td>
2382 <td>1</td>
2383 <td>0</td>
2384 </tr>
2385 <tr>
2386 <td>1</td>
2387 <td>0</td>
2388 <td>0</td>
2389 </tr>
2390 <tr>
2391 <td>1</td>
2392 <td>1</td>
2393 <td>1</td>
2394 </tr>
2395 </tbody>
2396</table>
2397</div>
2398<h5>Example:</h5>
2399<pre> &lt;result&gt; = and i32 4, %var <i>; yields {i32}:result = 4 &amp; %var</i>
2400 &lt;result&gt; = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2401 &lt;result&gt; = and i32 4, 8 <i>; yields {i32}:result = 0</i>
2402</pre>
2403</div>
2404<!-- _______________________________________________________________________ -->
2405<div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
2406<div class="doc_text">
2407<h5>Syntax:</h5>
2408<pre> &lt;result&gt; = or &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2409</pre>
2410<h5>Overview:</h5>
2411<p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2412or of its two operands.</p>
2413<h5>Arguments:</h5>
2414<p>The two arguments to the '<tt>or</tt>' instruction must be <a
2415 href="#t_integer">integer</a> values. Both arguments must have
2416identical types.</p>
2417<h5>Semantics:</h5>
2418<p>The truth table used for the '<tt>or</tt>' instruction is:</p>
2419<p> </p>
2420<div style="align: center">
2421<table border="1" cellspacing="0" cellpadding="4">
2422 <tbody>
2423 <tr>
2424 <td>In0</td>
2425 <td>In1</td>
2426 <td>Out</td>
2427 </tr>
2428 <tr>
2429 <td>0</td>
2430 <td>0</td>
2431 <td>0</td>
2432 </tr>
2433 <tr>
2434 <td>0</td>
2435 <td>1</td>
2436 <td>1</td>
2437 </tr>
2438 <tr>
2439 <td>1</td>
2440 <td>0</td>
2441 <td>1</td>
2442 </tr>
2443 <tr>
2444 <td>1</td>
2445 <td>1</td>
2446 <td>1</td>
2447 </tr>
2448 </tbody>
2449</table>
2450</div>
2451<h5>Example:</h5>
2452<pre> &lt;result&gt; = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2453 &lt;result&gt; = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2454 &lt;result&gt; = or i32 4, 8 <i>; yields {i32}:result = 12</i>
2455</pre>
2456</div>
2457<!-- _______________________________________________________________________ -->
2458<div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2459Instruction</a> </div>
2460<div class="doc_text">
2461<h5>Syntax:</h5>
2462<pre> &lt;result&gt; = xor &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {ty}:result</i>
2463</pre>
2464<h5>Overview:</h5>
2465<p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2466or of its two operands. The <tt>xor</tt> is used to implement the
2467"one's complement" operation, which is the "~" operator in C.</p>
2468<h5>Arguments:</h5>
2469<p>The two arguments to the '<tt>xor</tt>' instruction must be <a
2470 href="#t_integer">integer</a> values. Both arguments must have
2471identical types.</p>
2472<h5>Semantics:</h5>
2473<p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
2474<p> </p>
2475<div style="align: center">
2476<table border="1" cellspacing="0" cellpadding="4">
2477 <tbody>
2478 <tr>
2479 <td>In0</td>
2480 <td>In1</td>
2481 <td>Out</td>
2482 </tr>
2483 <tr>
2484 <td>0</td>
2485 <td>0</td>
2486 <td>0</td>
2487 </tr>
2488 <tr>
2489 <td>0</td>
2490 <td>1</td>
2491 <td>1</td>
2492 </tr>
2493 <tr>
2494 <td>1</td>
2495 <td>0</td>
2496 <td>1</td>
2497 </tr>
2498 <tr>
2499 <td>1</td>
2500 <td>1</td>
2501 <td>0</td>
2502 </tr>
2503 </tbody>
2504</table>
2505</div>
2506<p> </p>
2507<h5>Example:</h5>
2508<pre> &lt;result&gt; = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2509 &lt;result&gt; = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2510 &lt;result&gt; = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2511 &lt;result&gt; = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
2512</pre>
2513</div>
2514
2515<!-- ======================================================================= -->
2516<div class="doc_subsection">
2517 <a name="vectorops">Vector Operations</a>
2518</div>
2519
2520<div class="doc_text">
2521
2522<p>LLVM supports several instructions to represent vector operations in a
2523target-independent manner. These instructions cover the element-access and
2524vector-specific operations needed to process vectors effectively. While LLVM
2525does directly support these vector operations, many sophisticated algorithms
2526will want to use target-specific intrinsics to take full advantage of a specific
2527target.</p>
2528
2529</div>
2530
2531<!-- _______________________________________________________________________ -->
2532<div class="doc_subsubsection">
2533 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
2534</div>
2535
2536<div class="doc_text">
2537
2538<h5>Syntax:</h5>
2539
2540<pre>
2541 &lt;result&gt; = extractelement &lt;n x &lt;ty&gt;&gt; &lt;val&gt;, i32 &lt;idx&gt; <i>; yields &lt;ty&gt;</i>
2542</pre>
2543
2544<h5>Overview:</h5>
2545
2546<p>
2547The '<tt>extractelement</tt>' instruction extracts a single scalar
2548element from a vector at a specified index.
2549</p>
2550
2551
2552<h5>Arguments:</h5>
2553
2554<p>
2555The first operand of an '<tt>extractelement</tt>' instruction is a
2556value of <a href="#t_vector">vector</a> type. The second operand is
2557an index indicating the position from which to extract the element.
2558The index may be a variable.</p>
2559
2560<h5>Semantics:</h5>
2561
2562<p>
2563The result is a scalar of the same type as the element type of
2564<tt>val</tt>. Its value is the value at position <tt>idx</tt> of
2565<tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
2566results are undefined.
2567</p>
2568
2569<h5>Example:</h5>
2570
2571<pre>
2572 %result = extractelement &lt;4 x i32&gt; %vec, i32 0 <i>; yields i32</i>
2573</pre>
2574</div>
2575
2576
2577<!-- _______________________________________________________________________ -->
2578<div class="doc_subsubsection">
2579 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
2580</div>
2581
2582<div class="doc_text">
2583
2584<h5>Syntax:</h5>
2585
2586<pre>
2587 &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>
2588</pre>
2589
2590<h5>Overview:</h5>
2591
2592<p>
2593The '<tt>insertelement</tt>' instruction inserts a scalar
2594element into a vector at a specified index.
2595</p>
2596
2597
2598<h5>Arguments:</h5>
2599
2600<p>
2601The first operand of an '<tt>insertelement</tt>' instruction is a
2602value of <a href="#t_vector">vector</a> type. The second operand is a
2603scalar value whose type must equal the element type of the first
2604operand. The third operand is an index indicating the position at
2605which to insert the value. The index may be a variable.</p>
2606
2607<h5>Semantics:</h5>
2608
2609<p>
2610The result is a vector of the same type as <tt>val</tt>. Its
2611element values are those of <tt>val</tt> except at position
2612<tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
2613exceeds the length of <tt>val</tt>, the results are undefined.
2614</p>
2615
2616<h5>Example:</h5>
2617
2618<pre>
2619 %result = insertelement &lt;4 x i32&gt; %vec, i32 1, i32 0 <i>; yields &lt;4 x i32&gt;</i>
2620</pre>
2621</div>
2622
2623<!-- _______________________________________________________________________ -->
2624<div class="doc_subsubsection">
2625 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
2626</div>
2627
2628<div class="doc_text">
2629
2630<h5>Syntax:</h5>
2631
2632<pre>
2633 &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>
2634</pre>
2635
2636<h5>Overview:</h5>
2637
2638<p>
2639The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
2640from two input vectors, returning a vector of the same type.
2641</p>
2642
2643<h5>Arguments:</h5>
2644
2645<p>
2646The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
2647with types that match each other and types that match the result of the
2648instruction. The third argument is a shuffle mask, which has the same number
2649of elements as the other vector type, but whose element type is always 'i32'.
2650</p>
2651
2652<p>
2653The shuffle mask operand is required to be a constant vector with either
2654constant integer or undef values.
2655</p>
2656
2657<h5>Semantics:</h5>
2658
2659<p>
2660The elements of the two input vectors are numbered from left to right across
2661both of the vectors. The shuffle mask operand specifies, for each element of
2662the result vector, which element of the two input registers the result element
2663gets. The element selector may be undef (meaning "don't care") and the second
2664operand may be undef if performing a shuffle from only one vector.
2665</p>
2666
2667<h5>Example:</h5>
2668
2669<pre>
2670 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; %v2,
2671 &lt;4 x i32&gt; &lt;i32 0, i32 4, i32 1, i32 5&gt; <i>; yields &lt;4 x i32&gt;</i>
2672 %result = shufflevector &lt;4 x i32&gt; %v1, &lt;4 x i32&gt; undef,
2673 &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.
2674</pre>
2675</div>
2676
2677
2678<!-- ======================================================================= -->
2679<div class="doc_subsection">
2680 <a name="memoryops">Memory Access and Addressing Operations</a>
2681</div>
2682
2683<div class="doc_text">
2684
2685<p>A key design point of an SSA-based representation is how it
2686represents memory. In LLVM, no memory locations are in SSA form, which
2687makes things very simple. This section describes how to read, write,
2688allocate, and free memory in LLVM.</p>
2689
2690</div>
2691
2692<!-- _______________________________________________________________________ -->
2693<div class="doc_subsubsection">
2694 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
2695</div>
2696
2697<div class="doc_text">
2698
2699<h5>Syntax:</h5>
2700
2701<pre>
2702 &lt;result&gt; = malloc &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
2703</pre>
2704
2705<h5>Overview:</h5>
2706
2707<p>The '<tt>malloc</tt>' instruction allocates memory from the system
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]2708heap and returns a pointer to it. The object is always allocated in the generic
2709address space (address space zero).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2710
2711<h5>Arguments:</h5>
2712
2713<p>The '<tt>malloc</tt>' instruction allocates
2714<tt>sizeof(&lt;type&gt;)*NumElements</tt>
2715bytes of memory from the operating system and returns a pointer of the
2716appropriate type to the program. If "NumElements" is specified, it is the
2717number of elements allocated. If an alignment is specified, the value result
2718of the allocation is guaranteed to be aligned to at least that boundary. If
2719not specified, or if zero, the target can choose to align the allocation on any
2720convenient boundary.</p>
2721
2722<p>'<tt>type</tt>' must be a sized type.</p>
2723
2724<h5>Semantics:</h5>
2725
2726<p>Memory is allocated using the system "<tt>malloc</tt>" function, and
2727a pointer is returned.</p>
2728
2729<h5>Example:</h5>
2730
2731<pre>
2732 %array = malloc [4 x i8 ] <i>; yields {[%4 x i8]*}:array</i>
2733
2734 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
2735 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
2736 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
2737 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
2738 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
2739</pre>
2740</div>
2741
2742<!-- _______________________________________________________________________ -->
2743<div class="doc_subsubsection">
2744 <a name="i_free">'<tt>free</tt>' Instruction</a>
2745</div>
2746
2747<div class="doc_text">
2748
2749<h5>Syntax:</h5>
2750
2751<pre>
2752 free &lt;type&gt; &lt;value&gt; <i>; yields {void}</i>
2753</pre>
2754
2755<h5>Overview:</h5>
2756
2757<p>The '<tt>free</tt>' instruction returns memory back to the unused
2758memory heap to be reallocated in the future.</p>
2759
2760<h5>Arguments:</h5>
2761
2762<p>'<tt>value</tt>' shall be a pointer value that points to a value
2763that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
2764instruction.</p>
2765
2766<h5>Semantics:</h5>
2767
2768<p>Access to the memory pointed to by the pointer is no longer defined
2769after this instruction executes.</p>
2770
2771<h5>Example:</h5>
2772
2773<pre>
2774 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
2775 free [4 x i8]* %array
2776</pre>
2777</div>
2778
2779<!-- _______________________________________________________________________ -->
2780<div class="doc_subsubsection">
2781 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
2782</div>
2783
2784<div class="doc_text">
2785
2786<h5>Syntax:</h5>
2787
2788<pre>
2789 &lt;result&gt; = alloca &lt;type&gt;[, i32 &lt;NumElements&gt;][, align &lt;alignment&gt;] <i>; yields {type*}:result</i>
2790</pre>
2791
2792<h5>Overview:</h5>
2793
2794<p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
2795currently executing function, to be automatically released when this function
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]2796returns to its caller. The object is always allocated in the generic address
2797space (address space zero).</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]2798
2799<h5>Arguments:</h5>
2800
2801<p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(&lt;type&gt;)*NumElements</tt>
2802bytes of memory on the runtime stack, returning a pointer of the
2803appropriate type to the program. If "NumElements" is specified, it is the
2804number of elements allocated. If an alignment is specified, the value result
2805of the allocation is guaranteed to be aligned to at least that boundary. If
2806not specified, or if zero, the target can choose to align the allocation on any
2807convenient boundary.</p>
2808
2809<p>'<tt>type</tt>' may be any sized type.</p>
2810
2811<h5>Semantics:</h5>
2812
2813<p>Memory is allocated; a pointer is returned. '<tt>alloca</tt>'d
2814memory is automatically released when the function returns. The '<tt>alloca</tt>'
2815instruction is commonly used to represent automatic variables that must
2816have an address available. When the function returns (either with the <tt><a
2817 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
2818instructions), the memory is reclaimed.</p>
2819
2820<h5>Example:</h5>
2821
2822<pre>
2823 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
2824 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
2825 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
2826 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
2827</pre>
2828</div>
2829
2830<!-- _______________________________________________________________________ -->
2831<div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
2832Instruction</a> </div>
2833<div class="doc_text">
2834<h5>Syntax:</h5>
2835<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>
2836<h5>Overview:</h5>
2837<p>The '<tt>load</tt>' instruction is used to read from memory.</p>
2838<h5>Arguments:</h5>
2839<p>The argument to the '<tt>load</tt>' instruction specifies the memory
2840address from which to load. The pointer must point to a <a
2841 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
2842marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
2843the number or order of execution of this <tt>load</tt> with other
2844volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
2845instructions. </p>
2846<h5>Semantics:</h5>
2847<p>The location of memory pointed to is loaded.</p>
2848<h5>Examples:</h5>
2849<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
2850 <a
2851 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
2852 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
2853</pre>
2854</div>
2855<!-- _______________________________________________________________________ -->
2856<div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
2857Instruction</a> </div>
2858<div class="doc_text">
2859<h5>Syntax:</h5>
2860<pre> store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
2861 volatile store &lt;ty&gt; &lt;value&gt;, &lt;ty&gt;* &lt;pointer&gt;[, align &lt;alignment&gt;] <i>; yields {void}</i>
2862</pre>
2863<h5>Overview:</h5>
2864<p>The '<tt>store</tt>' instruction is used to write to memory.</p>
2865<h5>Arguments:</h5>
2866<p>There are two arguments to the '<tt>store</tt>' instruction: a value
2867to store and an address at which to store it. The type of the '<tt>&lt;pointer&gt;</tt>'
2868operand must be a pointer to the type of the '<tt>&lt;value&gt;</tt>'
2869operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
2870optimizer is not allowed to modify the number or order of execution of
2871this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
2872 href="#i_store">store</a></tt> instructions.</p>
2873<h5>Semantics:</h5>
2874<p>The contents of memory are updated to contain '<tt>&lt;value&gt;</tt>'
2875at the location specified by the '<tt>&lt;pointer&gt;</tt>' operand.</p>
2876<h5>Example:</h5>
2877<pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
Bill Wendling63ffa142007-10-22 05:10:05 +0000[diff] [blame]2878 store i32 3, i32* %ptr <i>; yields {void}</i>
2879 %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]2880</pre>
2881</div>
2882
2883<!-- _______________________________________________________________________ -->
2884<div class="doc_subsubsection">
2885 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
2886</div>
2887
2888<div class="doc_text">
2889<h5>Syntax:</h5>
2890<pre>
2891 &lt;result&gt; = getelementptr &lt;ty&gt;* &lt;ptrval&gt;{, &lt;ty&gt; &lt;idx&gt;}*
2892</pre>
2893
2894<h5>Overview:</h5>
2895
2896<p>
2897The '<tt>getelementptr</tt>' instruction is used to get the address of a
2898subelement of an aggregate data structure.</p>
2899
2900<h5>Arguments:</h5>
2901
2902<p>This instruction takes a list of integer operands that indicate what
2903elements of the aggregate object to index to. The actual types of the arguments
2904provided depend on the type of the first pointer argument. The
2905'<tt>getelementptr</tt>' instruction is used to index down through the type
2906levels of a structure or to a specific index in an array. When indexing into a
2907structure, only <tt>i32</tt> integer constants are allowed. When indexing
2908into an array or pointer, only integers of 32 or 64 bits are allowed, and will
2909be sign extended to 64-bit values.</p>
2910
2911<p>For example, let's consider a C code fragment and how it gets
2912compiled to LLVM:</p>
2913
2914<div class="doc_code">
2915<pre>
2916struct RT {
2917 char A;
2918 int B[10][20];
2919 char C;
2920};
2921struct ST {
2922 int X;
2923 double Y;
2924 struct RT Z;
2925};
2926
2927int *foo(struct ST *s) {
2928 return &amp;s[1].Z.B[5][13];
2929}
2930</pre>
2931</div>
2932
2933<p>The LLVM code generated by the GCC frontend is:</p>
2934
2935<div class="doc_code">
2936<pre>
2937%RT = type { i8 , [10 x [20 x i32]], i8 }
2938%ST = type { i32, double, %RT }
2939
2940define i32* %foo(%ST* %s) {
2941entry:
2942 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
2943 ret i32* %reg
2944}
2945</pre>
2946</div>
2947
2948<h5>Semantics:</h5>
2949
2950<p>The index types specified for the '<tt>getelementptr</tt>' instruction depend
2951on the pointer type that is being indexed into. <a href="#t_pointer">Pointer</a>
2952and <a href="#t_array">array</a> types can use a 32-bit or 64-bit
2953<a href="#t_integer">integer</a> type but the value will always be sign extended
2954to 64-bits. <a href="#t_struct">Structure</a> types require <tt>i32</tt>
2955<b>constants</b>.</p>
2956
2957<p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
2958type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
2959}</tt>' type, a structure. The second index indexes into the third element of
2960the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
2961i8 }</tt>' type, another structure. The third index indexes into the second
2962element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
2963array. The two dimensions of the array are subscripted into, yielding an
2964'<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
2965to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
2966
2967<p>Note that it is perfectly legal to index partially through a
2968structure, returning a pointer to an inner element. Because of this,
2969the LLVM code for the given testcase is equivalent to:</p>
2970
2971<pre>
2972 define i32* %foo(%ST* %s) {
2973 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
2974 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
2975 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
2976 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
2977 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
2978 ret i32* %t5
2979 }
2980</pre>
2981
2982<p>Note that it is undefined to access an array out of bounds: array and
2983pointer indexes must always be within the defined bounds of the array type.
2984The one exception for this rules is zero length arrays. These arrays are
2985defined to be accessible as variable length arrays, which requires access
2986beyond the zero'th element.</p>
2987
2988<p>The getelementptr instruction is often confusing. For some more insight
2989into how it works, see <a href="GetElementPtr.html">the getelementptr
2990FAQ</a>.</p>
2991
2992<h5>Example:</h5>
2993
2994<pre>
2995 <i>; yields [12 x i8]*:aptr</i>
2996 %aptr = getelementptr {i32, [12 x i8]}* %sptr, i64 0, i32 1
2997</pre>
2998</div>
2999
3000<!-- ======================================================================= -->
3001<div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
3002</div>
3003<div class="doc_text">
3004<p>The instructions in this category are the conversion instructions (casting)
3005which all take a single operand and a type. They perform various bit conversions
3006on the operand.</p>
3007</div>
3008
3009<!-- _______________________________________________________________________ -->
3010<div class="doc_subsubsection">
3011 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
3012</div>
3013<div class="doc_text">
3014
3015<h5>Syntax:</h5>
3016<pre>
3017 &lt;result&gt; = trunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3018</pre>
3019
3020<h5>Overview:</h5>
3021<p>
3022The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
3023</p>
3024
3025<h5>Arguments:</h5>
3026<p>
3027The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
3028be an <a href="#t_integer">integer</a> type, and a type that specifies the size
3029and type of the result, which must be an <a href="#t_integer">integer</a>
3030type. The bit size of <tt>value</tt> must be larger than the bit size of
3031<tt>ty2</tt>. Equal sized types are not allowed.</p>
3032
3033<h5>Semantics:</h5>
3034<p>
3035The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
3036and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
3037larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
3038It will always truncate bits.</p>
3039
3040<h5>Example:</h5>
3041<pre>
3042 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
3043 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
3044 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
3045</pre>
3046</div>
3047
3048<!-- _______________________________________________________________________ -->
3049<div class="doc_subsubsection">
3050 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
3051</div>
3052<div class="doc_text">
3053
3054<h5>Syntax:</h5>
3055<pre>
3056 &lt;result&gt; = zext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3057</pre>
3058
3059<h5>Overview:</h5>
3060<p>The '<tt>zext</tt>' instruction zero extends its operand to type
3061<tt>ty2</tt>.</p>
3062
3063
3064<h5>Arguments:</h5>
3065<p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
3066<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3067also be of <a href="#t_integer">integer</a> type. The bit size of the
3068<tt>value</tt> must be smaller than the bit size of the destination type,
3069<tt>ty2</tt>.</p>
3070
3071<h5>Semantics:</h5>
3072<p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
3073bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
3074
3075<p>When zero extending from i1, the result will always be either 0 or 1.</p>
3076
3077<h5>Example:</h5>
3078<pre>
3079 %X = zext i32 257 to i64 <i>; yields i64:257</i>
3080 %Y = zext i1 true to i32 <i>; yields i32:1</i>
3081</pre>
3082</div>
3083
3084<!-- _______________________________________________________________________ -->
3085<div class="doc_subsubsection">
3086 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
3087</div>
3088<div class="doc_text">
3089
3090<h5>Syntax:</h5>
3091<pre>
3092 &lt;result&gt; = sext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3093</pre>
3094
3095<h5>Overview:</h5>
3096<p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
3097
3098<h5>Arguments:</h5>
3099<p>
3100The '<tt>sext</tt>' instruction takes a value to cast, which must be of
3101<a href="#t_integer">integer</a> type, and a type to cast it to, which must
3102also be of <a href="#t_integer">integer</a> type. The bit size of the
3103<tt>value</tt> must be smaller than the bit size of the destination type,
3104<tt>ty2</tt>.</p>
3105
3106<h5>Semantics:</h5>
3107<p>
3108The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
3109bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
3110the type <tt>ty2</tt>.</p>
3111
3112<p>When sign extending from i1, the extension always results in -1 or 0.</p>
3113
3114<h5>Example:</h5>
3115<pre>
3116 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
3117 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
3118</pre>
3119</div>
3120
3121<!-- _______________________________________________________________________ -->
3122<div class="doc_subsubsection">
3123 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3124</div>
3125
3126<div class="doc_text">
3127
3128<h5>Syntax:</h5>
3129
3130<pre>
3131 &lt;result&gt; = fptrunc &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3132</pre>
3133
3134<h5>Overview:</h5>
3135<p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
3136<tt>ty2</tt>.</p>
3137
3138
3139<h5>Arguments:</h5>
3140<p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
3141 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
3142cast it to. The size of <tt>value</tt> must be larger than the size of
3143<tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3144<i>no-op cast</i>.</p>
3145
3146<h5>Semantics:</h5>
3147<p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3148<a href="#t_floating">floating point</a> type to a smaller
3149<a href="#t_floating">floating point</a> type. If the value cannot fit within
3150the destination type, <tt>ty2</tt>, then the results are undefined.</p>
3151
3152<h5>Example:</h5>
3153<pre>
3154 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3155 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3156</pre>
3157</div>
3158
3159<!-- _______________________________________________________________________ -->
3160<div class="doc_subsubsection">
3161 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
3162</div>
3163<div class="doc_text">
3164
3165<h5>Syntax:</h5>
3166<pre>
3167 &lt;result&gt; = fpext &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3168</pre>
3169
3170<h5>Overview:</h5>
3171<p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
3172floating point value.</p>
3173
3174<h5>Arguments:</h5>
3175<p>The '<tt>fpext</tt>' instruction takes a
3176<a href="#t_floating">floating point</a> <tt>value</tt> to cast,
3177and a <a href="#t_floating">floating point</a> type to cast it to. The source
3178type must be smaller than the destination type.</p>
3179
3180<h5>Semantics:</h5>
3181<p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
3182<a href="#t_floating">floating point</a> type to a larger
3183<a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
3184used to make a <i>no-op cast</i> because it always changes bits. Use
3185<tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
3186
3187<h5>Example:</h5>
3188<pre>
3189 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
3190 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
3191</pre>
3192</div>
3193
3194<!-- _______________________________________________________________________ -->
3195<div class="doc_subsubsection">
3196 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
3197</div>
3198<div class="doc_text">
3199
3200<h5>Syntax:</h5>
3201<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3202 &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]3203</pre>
3204
3205<h5>Overview:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3206<p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3207unsigned integer equivalent of type <tt>ty2</tt>.
3208</p>
3209
3210<h5>Arguments:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3211<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]3212scalar or vector <a href="#t_floating">floating point</a> value, and a type
3213to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3214type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3215vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3216
3217<h5>Semantics:</h5>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3218<p> The '<tt>fptoui</tt>' instruction converts its
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3219<a href="#t_floating">floating point</a> operand into the nearest (rounding
3220towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3221the results are undefined.</p>
3222
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3223<h5>Example:</h5>
3224<pre>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3225 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]3226 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
Reid Spencere6adee82007-07-31 14:40:14 +0000[diff] [blame]3227 %X = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3228</pre>
3229</div>
3230
3231<!-- _______________________________________________________________________ -->
3232<div class="doc_subsubsection">
3233 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
3234</div>
3235<div class="doc_text">
3236
3237<h5>Syntax:</h5>
3238<pre>
3239 &lt;result&gt; = fptosi &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3240</pre>
3241
3242<h5>Overview:</h5>
3243<p>The '<tt>fptosi</tt>' instruction converts
3244<a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
3245</p>
3246
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3247<h5>Arguments:</h5>
3248<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]3249scalar or vector <a href="#t_floating">floating point</a> value, and a type
3250to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3251type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3252vector integer type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3253
3254<h5>Semantics:</h5>
3255<p>The '<tt>fptosi</tt>' instruction converts its
3256<a href="#t_floating">floating point</a> operand into the nearest (rounding
3257towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3258the results are undefined.</p>
3259
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3260<h5>Example:</h5>
3261<pre>
3262 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
Chris Lattner681f1e82007-09-22 03:17:52 +0000[diff] [blame]3263 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3264 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
3265</pre>
3266</div>
3267
3268<!-- _______________________________________________________________________ -->
3269<div class="doc_subsubsection">
3270 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
3271</div>
3272<div class="doc_text">
3273
3274<h5>Syntax:</h5>
3275<pre>
3276 &lt;result&gt; = uitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3277</pre>
3278
3279<h5>Overview:</h5>
3280<p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
3281integer and converts that value to the <tt>ty2</tt> type.</p>
3282
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3283<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]3284<p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a
3285scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3286to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3287type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3288floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3289
3290<h5>Semantics:</h5>
3291<p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
3292integer quantity and converts it to the corresponding floating point value. If
3293the value cannot fit in the floating point value, the results are undefined.</p>
3294
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3295<h5>Example:</h5>
3296<pre>
3297 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
3298 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
3299</pre>
3300</div>
3301
3302<!-- _______________________________________________________________________ -->
3303<div class="doc_subsubsection">
3304 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
3305</div>
3306<div class="doc_text">
3307
3308<h5>Syntax:</h5>
3309<pre>
3310 &lt;result&gt; = sitofp &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3311</pre>
3312
3313<h5>Overview:</h5>
3314<p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
3315integer and converts that value to the <tt>ty2</tt> type.</p>
3316
3317<h5>Arguments:</h5>
Nate Begeman78246ca2007-11-17 03:58:34 +0000[diff] [blame]3318<p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a
3319scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3320to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3321type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3322floating point type with the same number of elements as <tt>ty</tt></p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3323
3324<h5>Semantics:</h5>
3325<p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
3326integer quantity and converts it to the corresponding floating point value. If
3327the value cannot fit in the floating point value, the results are undefined.</p>
3328
3329<h5>Example:</h5>
3330<pre>
3331 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
3332 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
3333</pre>
3334</div>
3335
3336<!-- _______________________________________________________________________ -->
3337<div class="doc_subsubsection">
3338 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3339</div>
3340<div class="doc_text">
3341
3342<h5>Syntax:</h5>
3343<pre>
3344 &lt;result&gt; = ptrtoint &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3345</pre>
3346
3347<h5>Overview:</h5>
3348<p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3349the integer type <tt>ty2</tt>.</p>
3350
3351<h5>Arguments:</h5>
3352<p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
3353must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
3354<tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.
3355
3356<h5>Semantics:</h5>
3357<p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3358<tt>ty2</tt> by interpreting the pointer value as an integer and either
3359truncating or zero extending that value to the size of the integer type. If
3360<tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3361<tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
3362are the same size, then nothing is done (<i>no-op cast</i>) other than a type
3363change.</p>
3364
3365<h5>Example:</h5>
3366<pre>
3367 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
3368 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
3369</pre>
3370</div>
3371
3372<!-- _______________________________________________________________________ -->
3373<div class="doc_subsubsection">
3374 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
3375</div>
3376<div class="doc_text">
3377
3378<h5>Syntax:</h5>
3379<pre>
3380 &lt;result&gt; = inttoptr &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3381</pre>
3382
3383<h5>Overview:</h5>
3384<p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
3385a pointer type, <tt>ty2</tt>.</p>
3386
3387<h5>Arguments:</h5>
3388<p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
3389value to cast, and a type to cast it to, which must be a
3390<a href="#t_pointer">pointer</a> type.
3391
3392<h5>Semantics:</h5>
3393<p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
3394<tt>ty2</tt> by applying either a zero extension or a truncation depending on
3395the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
3396size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
3397the size of a pointer then a zero extension is done. If they are the same size,
3398nothing is done (<i>no-op cast</i>).</p>
3399
3400<h5>Example:</h5>
3401<pre>
3402 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
3403 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
3404 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
3405</pre>
3406</div>
3407
3408<!-- _______________________________________________________________________ -->
3409<div class="doc_subsubsection">
3410 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
3411</div>
3412<div class="doc_text">
3413
3414<h5>Syntax:</h5>
3415<pre>
3416 &lt;result&gt; = bitcast &lt;ty&gt; &lt;value&gt; to &lt;ty2&gt; <i>; yields ty2</i>
3417</pre>
3418
3419<h5>Overview:</h5>
3420<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
3421<tt>ty2</tt> without changing any bits.</p>
3422
3423<h5>Arguments:</h5>
3424<p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
3425a first class value, and a type to cast it to, which must also be a <a
3426 href="#t_firstclass">first class</a> type. The bit sizes of <tt>value</tt>
3427and the destination type, <tt>ty2</tt>, must be identical. If the source
3428type is a pointer, the destination type must also be a pointer.</p>
3429
3430<h5>Semantics:</h5>
3431<p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
3432<tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
3433this conversion. The conversion is done as if the <tt>value</tt> had been
3434stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
3435converted to other pointer types with this instruction. To convert pointers to
3436other types, use the <a href="#i_inttoptr">inttoptr</a> or
3437<a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
3438
3439<h5>Example:</h5>
3440<pre>
3441 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
3442 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
3443 %Z = bitcast <2xint> %V to i64; <i>; yields i64: %V</i>
3444</pre>
3445</div>
3446
3447<!-- ======================================================================= -->
3448<div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
3449<div class="doc_text">
3450<p>The instructions in this category are the "miscellaneous"
3451instructions, which defy better classification.</p>
3452</div>
3453
3454<!-- _______________________________________________________________________ -->
3455<div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
3456</div>
3457<div class="doc_text">
3458<h5>Syntax:</h5>
3459<pre> &lt;result&gt; = icmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
3460</pre>
3461<h5>Overview:</h5>
3462<p>The '<tt>icmp</tt>' instruction returns a boolean value based on comparison
3463of its two integer operands.</p>
3464<h5>Arguments:</h5>
3465<p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
3466the condition code indicating the kind of comparison to perform. It is not
3467a value, just a keyword. The possible condition code are:
3468<ol>
3469 <li><tt>eq</tt>: equal</li>
3470 <li><tt>ne</tt>: not equal </li>
3471 <li><tt>ugt</tt>: unsigned greater than</li>
3472 <li><tt>uge</tt>: unsigned greater or equal</li>
3473 <li><tt>ult</tt>: unsigned less than</li>
3474 <li><tt>ule</tt>: unsigned less or equal</li>
3475 <li><tt>sgt</tt>: signed greater than</li>
3476 <li><tt>sge</tt>: signed greater or equal</li>
3477 <li><tt>slt</tt>: signed less than</li>
3478 <li><tt>sle</tt>: signed less or equal</li>
3479</ol>
3480<p>The remaining two arguments must be <a href="#t_integer">integer</a> or
3481<a href="#t_pointer">pointer</a> typed. They must also be identical types.</p>
3482<h5>Semantics:</h5>
3483<p>The '<tt>icmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3484the condition code given as <tt>cond</tt>. The comparison performed always
3485yields a <a href="#t_primitive">i1</a> result, as follows:
3486<ol>
3487 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
3488 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3489 </li>
3490 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
3491 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
3492 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
3493 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3494 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
3495 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3496 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
3497 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3498 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
3499 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3500 <li><tt>sgt</tt>: interprets the operands as signed values and yields
3501 <tt>true</tt> if <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3502 <li><tt>sge</tt>: interprets the operands as signed values and yields
3503 <tt>true</tt> if <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3504 <li><tt>slt</tt>: interprets the operands as signed values and yields
3505 <tt>true</tt> if <tt>var1</tt> is less than <tt>var2</tt>.</li>
3506 <li><tt>sle</tt>: interprets the operands as signed values and yields
3507 <tt>true</tt> if <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3508</ol>
3509<p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
3510values are compared as if they were integers.</p>
3511
3512<h5>Example:</h5>
3513<pre> &lt;result&gt; = icmp eq i32 4, 5 <i>; yields: result=false</i>
3514 &lt;result&gt; = icmp ne float* %X, %X <i>; yields: result=false</i>
3515 &lt;result&gt; = icmp ult i16 4, 5 <i>; yields: result=true</i>
3516 &lt;result&gt; = icmp sgt i16 4, 5 <i>; yields: result=false</i>
3517 &lt;result&gt; = icmp ule i16 -4, 5 <i>; yields: result=false</i>
3518 &lt;result&gt; = icmp sge i16 4, 5 <i>; yields: result=false</i>
3519</pre>
3520</div>
3521
3522<!-- _______________________________________________________________________ -->
3523<div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
3524</div>
3525<div class="doc_text">
3526<h5>Syntax:</h5>
3527<pre> &lt;result&gt; = fcmp &lt;cond&gt; &lt;ty&gt; &lt;var1&gt;, &lt;var2&gt; <i>; yields {i1}:result</i>
3528</pre>
3529<h5>Overview:</h5>
3530<p>The '<tt>fcmp</tt>' instruction returns a boolean value based on comparison
3531of its floating point operands.</p>
3532<h5>Arguments:</h5>
3533<p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
3534the condition code indicating the kind of comparison to perform. It is not
3535a value, just a keyword. The possible condition code are:
3536<ol>
3537 <li><tt>false</tt>: no comparison, always returns false</li>
3538 <li><tt>oeq</tt>: ordered and equal</li>
3539 <li><tt>ogt</tt>: ordered and greater than </li>
3540 <li><tt>oge</tt>: ordered and greater than or equal</li>
3541 <li><tt>olt</tt>: ordered and less than </li>
3542 <li><tt>ole</tt>: ordered and less than or equal</li>
3543 <li><tt>one</tt>: ordered and not equal</li>
3544 <li><tt>ord</tt>: ordered (no nans)</li>
3545 <li><tt>ueq</tt>: unordered or equal</li>
3546 <li><tt>ugt</tt>: unordered or greater than </li>
3547 <li><tt>uge</tt>: unordered or greater than or equal</li>
3548 <li><tt>ult</tt>: unordered or less than </li>
3549 <li><tt>ule</tt>: unordered or less than or equal</li>
3550 <li><tt>une</tt>: unordered or not equal</li>
3551 <li><tt>uno</tt>: unordered (either nans)</li>
3552 <li><tt>true</tt>: no comparison, always returns true</li>
3553</ol>
3554<p><i>Ordered</i> means that neither operand is a QNAN while
3555<i>unordered</i> means that either operand may be a QNAN.</p>
3556<p>The <tt>val1</tt> and <tt>val2</tt> arguments must be
3557<a href="#t_floating">floating point</a> typed. They must have identical
3558types.</p>
3559<h5>Semantics:</h5>
3560<p>The '<tt>fcmp</tt>' compares <tt>var1</tt> and <tt>var2</tt> according to
3561the condition code given as <tt>cond</tt>. The comparison performed always
3562yields a <a href="#t_primitive">i1</a> result, as follows:
3563<ol>
3564 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
3565 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3566 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
3567 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3568 <tt>var1</tt> is greather than <tt>var2</tt>.</li>
3569 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3570 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3571 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3572 <tt>var1</tt> is less than <tt>var2</tt>.</li>
3573 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3574 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3575 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
3576 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
3577 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
3578 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
3579 <tt>var1</tt> is equal to <tt>var2</tt>.</li>
3580 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
3581 <tt>var1</tt> is greater than <tt>var2</tt>.</li>
3582 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
3583 <tt>var1</tt> is greater than or equal to <tt>var2</tt>.</li>
3584 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
3585 <tt>var1</tt> is less than <tt>var2</tt>.</li>
3586 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
3587 <tt>var1</tt> is less than or equal to <tt>var2</tt>.</li>
3588 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
3589 <tt>var1</tt> is not equal to <tt>var2</tt>.</li>
3590 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
3591 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
3592</ol>
3593
3594<h5>Example:</h5>
3595<pre> &lt;result&gt; = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
3596 &lt;result&gt; = icmp one float 4.0, 5.0 <i>; yields: result=true</i>
3597 &lt;result&gt; = icmp olt float 4.0, 5.0 <i>; yields: result=true</i>
3598 &lt;result&gt; = icmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
3599</pre>
3600</div>
3601
3602<!-- _______________________________________________________________________ -->
3603<div class="doc_subsubsection"> <a name="i_phi">'<tt>phi</tt>'
3604Instruction</a> </div>
3605<div class="doc_text">
3606<h5>Syntax:</h5>
3607<pre> &lt;result&gt; = phi &lt;ty&gt; [ &lt;val0&gt;, &lt;label0&gt;], ...<br></pre>
3608<h5>Overview:</h5>
3609<p>The '<tt>phi</tt>' instruction is used to implement the &#966; node in
3610the SSA graph representing the function.</p>
3611<h5>Arguments:</h5>
3612<p>The type of the incoming values is specified with the first type
3613field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
3614as arguments, with one pair for each predecessor basic block of the
3615current block. Only values of <a href="#t_firstclass">first class</a>
3616type may be used as the value arguments to the PHI node. Only labels
3617may be used as the label arguments.</p>
3618<p>There must be no non-phi instructions between the start of a basic
3619block and the PHI instructions: i.e. PHI instructions must be first in
3620a basic block.</p>
3621<h5>Semantics:</h5>
3622<p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
3623specified by the pair corresponding to the predecessor basic block that executed
3624just prior to the current block.</p>
3625<h5>Example:</h5>
3626<pre>Loop: ; Infinite loop that counts from 0 on up...<br> %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]<br> %nextindvar = add i32 %indvar, 1<br> br label %Loop<br></pre>
3627</div>
3628
3629<!-- _______________________________________________________________________ -->
3630<div class="doc_subsubsection">
3631 <a name="i_select">'<tt>select</tt>' Instruction</a>
3632</div>
3633
3634<div class="doc_text">
3635
3636<h5>Syntax:</h5>
3637
3638<pre>
3639 &lt;result&gt; = select i1 &lt;cond&gt;, &lt;ty&gt; &lt;val1&gt;, &lt;ty&gt; &lt;val2&gt; <i>; yields ty</i>
3640</pre>
3641
3642<h5>Overview:</h5>
3643
3644<p>
3645The '<tt>select</tt>' instruction is used to choose one value based on a
3646condition, without branching.
3647</p>
3648
3649
3650<h5>Arguments:</h5>
3651
3652<p>
3653The '<tt>select</tt>' instruction requires a boolean value indicating the condition, and two values of the same <a href="#t_firstclass">first class</a> type.
3654</p>
3655
3656<h5>Semantics:</h5>
3657
3658<p>
3659If the boolean condition evaluates to true, the instruction returns the first
3660value argument; otherwise, it returns the second value argument.
3661</p>
3662
3663<h5>Example:</h5>
3664
3665<pre>
3666 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
3667</pre>
3668</div>
3669
3670
3671<!-- _______________________________________________________________________ -->
3672<div class="doc_subsubsection">
3673 <a name="i_call">'<tt>call</tt>' Instruction</a>
3674</div>
3675
3676<div class="doc_text">
3677
3678<h5>Syntax:</h5>
3679<pre>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]3680 &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]3681</pre>
3682
3683<h5>Overview:</h5>
3684
3685<p>The '<tt>call</tt>' instruction represents a simple function call.</p>
3686
3687<h5>Arguments:</h5>
3688
3689<p>This instruction requires several arguments:</p>
3690
3691<ol>
3692 <li>
3693 <p>The optional "tail" marker indicates whether the callee function accesses
3694 any allocas or varargs in the caller. If the "tail" marker is present, the
3695 function call is eligible for tail call optimization. Note that calls may
3696 be marked "tail" even if they do not occur before a <a
3697 href="#i_ret"><tt>ret</tt></a> instruction.
3698 </li>
3699 <li>
3700 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
3701 convention</a> the call should use. If none is specified, the call defaults
3702 to using C calling conventions.
3703 </li>
3704 <li>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]3705 <p>'<tt>ty</tt>': the type of the call instruction itself which is also
3706 the type of the return value. Functions that return no value are marked
3707 <tt><a href="#t_void">void</a></tt>.</p>
3708 </li>
3709 <li>
3710 <p>'<tt>fnty</tt>': shall be the signature of the pointer to function
3711 value being invoked. The argument types must match the types implied by
3712 this signature. This type can be omitted if the function is not varargs
3713 and if the function type does not return a pointer to a function.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3714 </li>
3715 <li>
3716 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
3717 be invoked. In most cases, this is a direct function invocation, but
3718 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
3719 to function value.</p>
3720 </li>
3721 <li>
3722 <p>'<tt>function args</tt>': argument list whose types match the
3723 function signature argument types. All arguments must be of
3724 <a href="#t_firstclass">first class</a> type. If the function signature
3725 indicates the function accepts a variable number of arguments, the extra
3726 arguments can be specified.</p>
3727 </li>
3728</ol>
3729
3730<h5>Semantics:</h5>
3731
3732<p>The '<tt>call</tt>' instruction is used to cause control flow to
3733transfer to a specified function, with its incoming arguments bound to
3734the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
3735instruction in the called function, control flow continues with the
3736instruction after the function call, and the return value of the
3737function is bound to the result argument. This is a simpler case of
3738the <a href="#i_invoke">invoke</a> instruction.</p>
3739
3740<h5>Example:</h5>
3741
3742<pre>
Nick Lewycky93082fc2007-09-08 13:57:50 +0000[diff] [blame]3743 %retval = call i32 @test(i32 %argc)
3744 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42);
3745 %X = tail call i32 @foo()
3746 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo()
3747 %Z = call void %foo(i8 97 signext)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3748</pre>
3749
3750</div>
3751
3752<!-- _______________________________________________________________________ -->
3753<div class="doc_subsubsection">
3754 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
3755</div>
3756
3757<div class="doc_text">
3758
3759<h5>Syntax:</h5>
3760
3761<pre>
3762 &lt;resultval&gt; = va_arg &lt;va_list*&gt; &lt;arglist&gt;, &lt;argty&gt;
3763</pre>
3764
3765<h5>Overview:</h5>
3766
3767<p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
3768the "variable argument" area of a function call. It is used to implement the
3769<tt>va_arg</tt> macro in C.</p>
3770
3771<h5>Arguments:</h5>
3772
3773<p>This instruction takes a <tt>va_list*</tt> value and the type of
3774the argument. It returns a value of the specified argument type and
3775increments the <tt>va_list</tt> to point to the next argument. The
3776actual type of <tt>va_list</tt> is target specific.</p>
3777
3778<h5>Semantics:</h5>
3779
3780<p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
3781type from the specified <tt>va_list</tt> and causes the
3782<tt>va_list</tt> to point to the next argument. For more information,
3783see the variable argument handling <a href="#int_varargs">Intrinsic
3784Functions</a>.</p>
3785
3786<p>It is legal for this instruction to be called in a function which does not
3787take a variable number of arguments, for example, the <tt>vfprintf</tt>
3788function.</p>
3789
3790<p><tt>va_arg</tt> is an LLVM instruction instead of an <a
3791href="#intrinsics">intrinsic function</a> because it takes a type as an
3792argument.</p>
3793
3794<h5>Example:</h5>
3795
3796<p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
3797
3798</div>
3799
3800<!-- *********************************************************************** -->
3801<div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
3802<!-- *********************************************************************** -->
3803
3804<div class="doc_text">
3805
3806<p>LLVM supports the notion of an "intrinsic function". These functions have
3807well known names and semantics and are required to follow certain restrictions.
3808Overall, these intrinsics represent an extension mechanism for the LLVM
3809language that does not require changing all of the transformations in LLVM when
3810adding to the language (or the bitcode reader/writer, the parser, etc...).</p>
3811
3812<p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
3813prefix is reserved in LLVM for intrinsic names; thus, function names may not
3814begin with this prefix. Intrinsic functions must always be external functions:
3815you cannot define the body of intrinsic functions. Intrinsic functions may
3816only be used in call or invoke instructions: it is illegal to take the address
3817of an intrinsic function. Additionally, because intrinsic functions are part
3818of the LLVM language, it is required if any are added that they be documented
3819here.</p>
3820
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]3821<p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
3822a family of functions that perform the same operation but on different data
3823types. Because LLVM can represent over 8 million different integer types,
3824overloading is used commonly to allow an intrinsic function to operate on any
3825integer type. One or more of the argument types or the result type can be
3826overloaded to accept any integer type. Argument types may also be defined as
3827exactly matching a previous argument's type or the result type. This allows an
3828intrinsic function which accepts multiple arguments, but needs all of them to
3829be of the same type, to only be overloaded with respect to a single argument or
3830the result.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3831
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]3832<p>Overloaded intrinsics will have the names of its overloaded argument types
3833encoded into its function name, each preceded by a period. Only those types
3834which are overloaded result in a name suffix. Arguments whose type is matched
3835against another type do not. For example, the <tt>llvm.ctpop</tt> function can
3836take an integer of any width and returns an integer of exactly the same integer
3837width. This leads to a family of functions such as
3838<tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29 %val)</tt>.
3839Only one type, the return type, is overloaded, and only one type suffix is
3840required. Because the argument's type is matched against the return type, it
3841does not require its own name suffix.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]3842
3843<p>To learn how to add an intrinsic function, please see the
3844<a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
3845</p>
3846
3847</div>
3848
3849<!-- ======================================================================= -->
3850<div class="doc_subsection">
3851 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
3852</div>
3853
3854<div class="doc_text">
3855
3856<p>Variable argument support is defined in LLVM with the <a
3857 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
3858intrinsic functions. These functions are related to the similarly
3859named macros defined in the <tt>&lt;stdarg.h&gt;</tt> header file.</p>
3860
3861<p>All of these functions operate on arguments that use a
3862target-specific value type "<tt>va_list</tt>". The LLVM assembly
3863language reference manual does not define what this type is, so all
3864transformations should be prepared to handle these functions regardless of
3865the type used.</p>
3866
3867<p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
3868instruction and the variable argument handling intrinsic functions are
3869used.</p>
3870
3871<div class="doc_code">
3872<pre>
3873define i32 @test(i32 %X, ...) {
3874 ; Initialize variable argument processing
3875 %ap = alloca i8*
3876 %ap2 = bitcast i8** %ap to i8*
3877 call void @llvm.va_start(i8* %ap2)
3878
3879 ; Read a single integer argument
3880 %tmp = va_arg i8** %ap, i32
3881
3882 ; Demonstrate usage of llvm.va_copy and llvm.va_end
3883 %aq = alloca i8*
3884 %aq2 = bitcast i8** %aq to i8*
3885 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
3886 call void @llvm.va_end(i8* %aq2)
3887
3888 ; Stop processing of arguments.
3889 call void @llvm.va_end(i8* %ap2)
3890 ret i32 %tmp
3891}
3892
3893declare void @llvm.va_start(i8*)
3894declare void @llvm.va_copy(i8*, i8*)
3895declare void @llvm.va_end(i8*)
3896</pre>
3897</div>
3898
3899</div>
3900
3901<!-- _______________________________________________________________________ -->
3902<div class="doc_subsubsection">
3903 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
3904</div>
3905
3906
3907<div class="doc_text">
3908<h5>Syntax:</h5>
3909<pre> declare void %llvm.va_start(i8* &lt;arglist&gt;)<br></pre>
3910<h5>Overview:</h5>
3911<P>The '<tt>llvm.va_start</tt>' intrinsic initializes
3912<tt>*&lt;arglist&gt;</tt> for subsequent use by <tt><a
3913href="#i_va_arg">va_arg</a></tt>.</p>
3914
3915<h5>Arguments:</h5>
3916
3917<P>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
3918
3919<h5>Semantics:</h5>
3920
3921<P>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
3922macro available in C. In a target-dependent way, it initializes the
3923<tt>va_list</tt> element to which the argument points, so that the next call to
3924<tt>va_arg</tt> will produce the first variable argument passed to the function.
3925Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
3926last argument of the function as the compiler can figure that out.</p>
3927
3928</div>
3929
3930<!-- _______________________________________________________________________ -->
3931<div class="doc_subsubsection">
3932 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
3933</div>
3934
3935<div class="doc_text">
3936<h5>Syntax:</h5>
3937<pre> declare void @llvm.va_end(i8* &lt;arglist&gt;)<br></pre>
3938<h5>Overview:</h5>
3939
3940<p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*&lt;arglist&gt;</tt>,
3941which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
3942or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
3943
3944<h5>Arguments:</h5>
3945
3946<p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
3947
3948<h5>Semantics:</h5>
3949
3950<p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
3951macro available in C. In a target-dependent way, it destroys the
3952<tt>va_list</tt> element to which the argument points. Calls to <a
3953href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
3954<tt>llvm.va_copy</tt></a> must be matched exactly with calls to
3955<tt>llvm.va_end</tt>.</p>
3956
3957</div>
3958
3959<!-- _______________________________________________________________________ -->
3960<div class="doc_subsubsection">
3961 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
3962</div>
3963
3964<div class="doc_text">
3965
3966<h5>Syntax:</h5>
3967
3968<pre>
3969 declare void @llvm.va_copy(i8* &lt;destarglist&gt;, i8* &lt;srcarglist&gt;)
3970</pre>
3971
3972<h5>Overview:</h5>
3973
3974<p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
3975from the source argument list to the destination argument list.</p>
3976
3977<h5>Arguments:</h5>
3978
3979<p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
3980The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
3981
3982
3983<h5>Semantics:</h5>
3984
3985<p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
3986macro available in C. In a target-dependent way, it copies the source
3987<tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
3988intrinsic is necessary because the <tt><a href="#int_va_start">
3989llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
3990example, memory allocation.</p>
3991
3992</div>
3993
3994<!-- ======================================================================= -->
3995<div class="doc_subsection">
3996 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
3997</div>
3998
3999<div class="doc_text">
4000
4001<p>
4002LLVM support for <a href="GarbageCollection.html">Accurate Garbage
4003Collection</a> requires the implementation and generation of these intrinsics.
4004These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
4005stack</a>, as well as garbage collector implementations that require <a
4006href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
4007Front-ends for type-safe garbage collected languages should generate these
4008intrinsics to make use of the LLVM garbage collectors. For more details, see <a
4009href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
4010</p>
Christopher Lambcfe00962007-12-17 01:00:21 +0000[diff] [blame]4011
4012<p>The garbage collection intrinsics only operate on objects in the generic
4013 address space (address space zero).</p>
4014
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4015</div>
4016
4017<!-- _______________________________________________________________________ -->
4018<div class="doc_subsubsection">
4019 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
4020</div>
4021
4022<div class="doc_text">
4023
4024<h5>Syntax:</h5>
4025
4026<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4027 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4028</pre>
4029
4030<h5>Overview:</h5>
4031
4032<p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
4033the code generator, and allows some metadata to be associated with it.</p>
4034
4035<h5>Arguments:</h5>
4036
4037<p>The first argument specifies the address of a stack object that contains the
4038root pointer. The second pointer (which must be either a constant or a global
4039value address) contains the meta-data to be associated with the root.</p>
4040
4041<h5>Semantics:</h5>
4042
4043<p>At runtime, a call to this intrinsics stores a null pointer into the "ptrloc"
4044location. At compile-time, the code generator generates information to allow
Gordon Henriksen40393542007-12-25 02:31:26 +0000[diff] [blame]4045the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
4046intrinsic may only be used in a function which <a href="#gc">specifies a GC
4047algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4048
4049</div>
4050
4051
4052<!-- _______________________________________________________________________ -->
4053<div class="doc_subsubsection">
4054 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
4055</div>
4056
4057<div class="doc_text">
4058
4059<h5>Syntax:</h5>
4060
4061<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4062 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4063</pre>
4064
4065<h5>Overview:</h5>
4066
4067<p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
4068locations, allowing garbage collector implementations that require read
4069barriers.</p>
4070
4071<h5>Arguments:</h5>
4072
4073<p>The second argument is the address to read from, which should be an address
4074allocated from the garbage collector. The first object is a pointer to the
4075start of the referenced object, if needed by the language runtime (otherwise
4076null).</p>
4077
4078<h5>Semantics:</h5>
4079
4080<p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
4081instruction, but may be replaced with substantially more complex code by the
Gordon Henriksen40393542007-12-25 02:31:26 +0000[diff] [blame]4082garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
4083may only be used in a function which <a href="#gc">specifies a GC
4084algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4085
4086</div>
4087
4088
4089<!-- _______________________________________________________________________ -->
4090<div class="doc_subsubsection">
4091 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
4092</div>
4093
4094<div class="doc_text">
4095
4096<h5>Syntax:</h5>
4097
4098<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4099 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4100</pre>
4101
4102<h5>Overview:</h5>
4103
4104<p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
4105locations, allowing garbage collector implementations that require write
4106barriers (such as generational or reference counting collectors).</p>
4107
4108<h5>Arguments:</h5>
4109
4110<p>The first argument is the reference to store, the second is the start of the
4111object to store it to, and the third is the address of the field of Obj to
4112store to. If the runtime does not require a pointer to the object, Obj may be
4113null.</p>
4114
4115<h5>Semantics:</h5>
4116
4117<p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
4118instruction, but may be replaced with substantially more complex code by the
Gordon Henriksen40393542007-12-25 02:31:26 +0000[diff] [blame]4119garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
4120may only be used in a function which <a href="#gc">specifies a GC
4121algorithm</a>.</p>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4122
4123</div>
4124
4125
4126
4127<!-- ======================================================================= -->
4128<div class="doc_subsection">
4129 <a name="int_codegen">Code Generator Intrinsics</a>
4130</div>
4131
4132<div class="doc_text">
4133<p>
4134These intrinsics are provided by LLVM to expose special features that may only
4135be implemented with code generator support.
4136</p>
4137
4138</div>
4139
4140<!-- _______________________________________________________________________ -->
4141<div class="doc_subsubsection">
4142 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
4143</div>
4144
4145<div class="doc_text">
4146
4147<h5>Syntax:</h5>
4148<pre>
4149 declare i8 *@llvm.returnaddress(i32 &lt;level&gt;)
4150</pre>
4151
4152<h5>Overview:</h5>
4153
4154<p>
4155The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
4156target-specific value indicating the return address of the current function
4157or one of its callers.
4158</p>
4159
4160<h5>Arguments:</h5>
4161
4162<p>
4163The argument to this intrinsic indicates which function to return the address
4164for. Zero indicates the calling function, one indicates its caller, etc. The
4165argument is <b>required</b> to be a constant integer value.
4166</p>
4167
4168<h5>Semantics:</h5>
4169
4170<p>
4171The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
4172the return address of the specified call frame, or zero if it cannot be
4173identified. The value returned by this intrinsic is likely to be incorrect or 0
4174for arguments other than zero, so it should only be used for debugging purposes.
4175</p>
4176
4177<p>
4178Note that calling this intrinsic does not prevent function inlining or other
4179aggressive transformations, so the value returned may not be that of the obvious
4180source-language caller.
4181</p>
4182</div>
4183
4184
4185<!-- _______________________________________________________________________ -->
4186<div class="doc_subsubsection">
4187 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
4188</div>
4189
4190<div class="doc_text">
4191
4192<h5>Syntax:</h5>
4193<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4194 declare i8 *@llvm.frameaddress(i32 &lt;level&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4195</pre>
4196
4197<h5>Overview:</h5>
4198
4199<p>
4200The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
4201target-specific frame pointer value for the specified stack frame.
4202</p>
4203
4204<h5>Arguments:</h5>
4205
4206<p>
4207The argument to this intrinsic indicates which function to return the frame
4208pointer for. Zero indicates the calling function, one indicates its caller,
4209etc. The argument is <b>required</b> to be a constant integer value.
4210</p>
4211
4212<h5>Semantics:</h5>
4213
4214<p>
4215The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
4216the frame address of the specified call frame, or zero if it cannot be
4217identified. The value returned by this intrinsic is likely to be incorrect or 0
4218for arguments other than zero, so it should only be used for debugging purposes.
4219</p>
4220
4221<p>
4222Note that calling this intrinsic does not prevent function inlining or other
4223aggressive transformations, so the value returned may not be that of the obvious
4224source-language caller.
4225</p>
4226</div>
4227
4228<!-- _______________________________________________________________________ -->
4229<div class="doc_subsubsection">
4230 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
4231</div>
4232
4233<div class="doc_text">
4234
4235<h5>Syntax:</h5>
4236<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4237 declare i8 *@llvm.stacksave()
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4238</pre>
4239
4240<h5>Overview:</h5>
4241
4242<p>
4243The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
4244the function stack, for use with <a href="#int_stackrestore">
4245<tt>llvm.stackrestore</tt></a>. This is useful for implementing language
4246features like scoped automatic variable sized arrays in C99.
4247</p>
4248
4249<h5>Semantics:</h5>
4250
4251<p>
4252This intrinsic returns a opaque pointer value that can be passed to <a
4253href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
4254<tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
4255<tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
4256state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
4257practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
4258that were allocated after the <tt>llvm.stacksave</tt> was executed.
4259</p>
4260
4261</div>
4262
4263<!-- _______________________________________________________________________ -->
4264<div class="doc_subsubsection">
4265 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
4266</div>
4267
4268<div class="doc_text">
4269
4270<h5>Syntax:</h5>
4271<pre>
4272 declare void @llvm.stackrestore(i8 * %ptr)
4273</pre>
4274
4275<h5>Overview:</h5>
4276
4277<p>
4278The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
4279the function stack to the state it was in when the corresponding <a
4280href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
4281useful for implementing language features like scoped automatic variable sized
4282arrays in C99.
4283</p>
4284
4285<h5>Semantics:</h5>
4286
4287<p>
4288See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
4289</p>
4290
4291</div>
4292
4293
4294<!-- _______________________________________________________________________ -->
4295<div class="doc_subsubsection">
4296 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
4297</div>
4298
4299<div class="doc_text">
4300
4301<h5>Syntax:</h5>
4302<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4303 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]4304</pre>
4305
4306<h5>Overview:</h5>
4307
4308
4309<p>
4310The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
4311a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
4312no
4313effect on the behavior of the program but can change its performance
4314characteristics.
4315</p>
4316
4317<h5>Arguments:</h5>
4318
4319<p>
4320<tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
4321determining if the fetch should be for a read (0) or write (1), and
4322<tt>locality</tt> is a temporal locality specifier ranging from (0) - no
4323locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
4324<tt>locality</tt> arguments must be constant integers.
4325</p>
4326
4327<h5>Semantics:</h5>
4328
4329<p>
4330This intrinsic does not modify the behavior of the program. In particular,
4331prefetches cannot trap and do not produce a value. On targets that support this
4332intrinsic, the prefetch can provide hints to the processor cache for better
4333performance.
4334</p>
4335
4336</div>
4337
4338<!-- _______________________________________________________________________ -->
4339<div class="doc_subsubsection">
4340 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
4341</div>
4342
4343<div class="doc_text">
4344
4345<h5>Syntax:</h5>
4346<pre>
Chris Lattner38bd5dd2007-09-21 17:30:40 +0000[diff] [blame]4347 declare void @llvm.pcmarker(i32 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4348</pre>
4349
4350<h5>Overview:</h5>
4351
4352
4353<p>
4354The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
4355(PC) in a region of
4356code to simulators and other tools. The method is target specific, but it is
4357expected that the marker will use exported symbols to transmit the PC of the marker.
4358The marker makes no guarantees that it will remain with any specific instruction
4359after optimizations. It is possible that the presence of a marker will inhibit
4360optimizations. The intended use is to be inserted after optimizations to allow
4361correlations of simulation runs.
4362</p>
4363
4364<h5>Arguments:</h5>
4365
4366<p>
4367<tt>id</tt> is a numerical id identifying the marker.
4368</p>
4369
4370<h5>Semantics:</h5>
4371
4372<p>
4373This intrinsic does not modify the behavior of the program. Backends that do not
4374support this intrinisic may ignore it.
4375</p>
4376
4377</div>
4378
4379<!-- _______________________________________________________________________ -->
4380<div class="doc_subsubsection">
4381 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
4382</div>
4383
4384<div class="doc_text">
4385
4386<h5>Syntax:</h5>
4387<pre>
4388 declare i64 @llvm.readcyclecounter( )
4389</pre>
4390
4391<h5>Overview:</h5>
4392
4393
4394<p>
4395The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
4396counter register (or similar low latency, high accuracy clocks) on those targets
4397that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
4398As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
4399should only be used for small timings.
4400</p>
4401
4402<h5>Semantics:</h5>
4403
4404<p>
4405When directly supported, reading the cycle counter should not modify any memory.
4406Implementations are allowed to either return a application specific value or a
4407system wide value. On backends without support, this is lowered to a constant 0.
4408</p>
4409
4410</div>
4411
4412<!-- ======================================================================= -->
4413<div class="doc_subsection">
4414 <a name="int_libc">Standard C Library Intrinsics</a>
4415</div>
4416
4417<div class="doc_text">
4418<p>
4419LLVM provides intrinsics for a few important standard C library functions.
4420These intrinsics allow source-language front-ends to pass information about the
4421alignment of the pointer arguments to the code generator, providing opportunity
4422for more efficient code generation.
4423</p>
4424
4425</div>
4426
4427<!-- _______________________________________________________________________ -->
4428<div class="doc_subsubsection">
4429 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
4430</div>
4431
4432<div class="doc_text">
4433
4434<h5>Syntax:</h5>
4435<pre>
4436 declare void @llvm.memcpy.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4437 i32 &lt;len&gt;, i32 &lt;align&gt;)
4438 declare void @llvm.memcpy.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4439 i64 &lt;len&gt;, i32 &lt;align&gt;)
4440</pre>
4441
4442<h5>Overview:</h5>
4443
4444<p>
4445The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
4446location to the destination location.
4447</p>
4448
4449<p>
4450Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
4451intrinsics do not return a value, and takes an extra alignment argument.
4452</p>
4453
4454<h5>Arguments:</h5>
4455
4456<p>
4457The first argument is a pointer to the destination, the second is a pointer to
4458the source. The third argument is an integer argument
4459specifying the number of bytes to copy, and the fourth argument is the alignment
4460of the source and destination locations.
4461</p>
4462
4463<p>
4464If the call to this intrinisic has an alignment value that is not 0 or 1, then
4465the caller guarantees that both the source and destination pointers are aligned
4466to that boundary.
4467</p>
4468
4469<h5>Semantics:</h5>
4470
4471<p>
4472The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
4473location to the destination location, which are not allowed to overlap. It
4474copies "len" bytes of memory over. If the argument is known to be aligned to
4475some boundary, this can be specified as the fourth argument, otherwise it should
4476be set to 0 or 1.
4477</p>
4478</div>
4479
4480
4481<!-- _______________________________________________________________________ -->
4482<div class="doc_subsubsection">
4483 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
4484</div>
4485
4486<div class="doc_text">
4487
4488<h5>Syntax:</h5>
4489<pre>
4490 declare void @llvm.memmove.i32(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4491 i32 &lt;len&gt;, i32 &lt;align&gt;)
4492 declare void @llvm.memmove.i64(i8 * &lt;dest&gt;, i8 * &lt;src&gt;,
4493 i64 &lt;len&gt;, i32 &lt;align&gt;)
4494</pre>
4495
4496<h5>Overview:</h5>
4497
4498<p>
4499The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
4500location to the destination location. It is similar to the
4501'<tt>llvm.memcmp</tt>' intrinsic but allows the two memory locations to overlap.
4502</p>
4503
4504<p>
4505Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
4506intrinsics do not return a value, and takes an extra alignment argument.
4507</p>
4508
4509<h5>Arguments:</h5>
4510
4511<p>
4512The first argument is a pointer to the destination, the second is a pointer to
4513the source. The third argument is an integer argument
4514specifying the number of bytes to copy, and the fourth argument is the alignment
4515of the source and destination locations.
4516</p>
4517
4518<p>
4519If the call to this intrinisic has an alignment value that is not 0 or 1, then
4520the caller guarantees that the source and destination pointers are aligned to
4521that boundary.
4522</p>
4523
4524<h5>Semantics:</h5>
4525
4526<p>
4527The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
4528location to the destination location, which may overlap. It
4529copies "len" bytes of memory over. If the argument is known to be aligned to
4530some boundary, this can be specified as the fourth argument, otherwise it should
4531be set to 0 or 1.
4532</p>
4533</div>
4534
4535
4536<!-- _______________________________________________________________________ -->
4537<div class="doc_subsubsection">
4538 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
4539</div>
4540
4541<div class="doc_text">
4542
4543<h5>Syntax:</h5>
4544<pre>
4545 declare void @llvm.memset.i32(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
4546 i32 &lt;len&gt;, i32 &lt;align&gt;)
4547 declare void @llvm.memset.i64(i8 * &lt;dest&gt;, i8 &lt;val&gt;,
4548 i64 &lt;len&gt;, i32 &lt;align&gt;)
4549</pre>
4550
4551<h5>Overview:</h5>
4552
4553<p>
4554The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
4555byte value.
4556</p>
4557
4558<p>
4559Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
4560does not return a value, and takes an extra alignment argument.
4561</p>
4562
4563<h5>Arguments:</h5>
4564
4565<p>
4566The first argument is a pointer to the destination to fill, the second is the
4567byte value to fill it with, the third argument is an integer
4568argument specifying the number of bytes to fill, and the fourth argument is the
4569known alignment of destination location.
4570</p>
4571
4572<p>
4573If the call to this intrinisic has an alignment value that is not 0 or 1, then
4574the caller guarantees that the destination pointer is aligned to that boundary.
4575</p>
4576
4577<h5>Semantics:</h5>
4578
4579<p>
4580The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
4581the
4582destination location. If the argument is known to be aligned to some boundary,
4583this can be specified as the fourth argument, otherwise it should be set to 0 or
45841.
4585</p>
4586</div>
4587
4588
4589<!-- _______________________________________________________________________ -->
4590<div class="doc_subsubsection">
4591 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
4592</div>
4593
4594<div class="doc_text">
4595
4596<h5>Syntax:</h5>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]4597<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]4598floating point or vector of floating point type. Not all targets support all
4599types however.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4600<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]4601 declare float @llvm.sqrt.f32(float %Val)
4602 declare double @llvm.sqrt.f64(double %Val)
4603 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
4604 declare fp128 @llvm.sqrt.f128(fp128 %Val)
4605 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4606</pre>
4607
4608<h5>Overview:</h5>
4609
4610<p>
4611The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]4612returning the same value as the libm '<tt>sqrt</tt>' functions would. Unlike
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4613<tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
4614negative numbers (which allows for better optimization).
4615</p>
4616
4617<h5>Arguments:</h5>
4618
4619<p>
4620The argument and return value are floating point numbers of the same type.
4621</p>
4622
4623<h5>Semantics:</h5>
4624
4625<p>
4626This function returns the sqrt of the specified operand if it is a nonnegative
4627floating point number.
4628</p>
4629</div>
4630
4631<!-- _______________________________________________________________________ -->
4632<div class="doc_subsubsection">
4633 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
4634</div>
4635
4636<div class="doc_text">
4637
4638<h5>Syntax:</h5>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]4639<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]4640floating point or vector of floating point type. Not all targets support all
4641types however.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4642<pre>
Dale Johannesenf9adbb62007-10-02 17:47:38 +0000[diff] [blame]4643 declare float @llvm.powi.f32(float %Val, i32 %power)
4644 declare double @llvm.powi.f64(double %Val, i32 %power)
4645 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
4646 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
4647 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4648</pre>
4649
4650<h5>Overview:</h5>
4651
4652<p>
4653The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
4654specified (positive or negative) power. The order of evaluation of
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]4655multiplications is not defined. When a vector of floating point type is
4656used, the second argument remains a scalar integer value.
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4657</p>
4658
4659<h5>Arguments:</h5>
4660
4661<p>
4662The second argument is an integer power, and the first is a value to raise to
4663that power.
4664</p>
4665
4666<h5>Semantics:</h5>
4667
4668<p>
4669This function returns the first value raised to the second power with an
4670unspecified sequence of rounding operations.</p>
4671</div>
4672
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]4673<!-- _______________________________________________________________________ -->
4674<div class="doc_subsubsection">
4675 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
4676</div>
4677
4678<div class="doc_text">
4679
4680<h5>Syntax:</h5>
4681<p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
4682floating point or vector of floating point type. Not all targets support all
4683types however.
4684<pre>
4685 declare float @llvm.sin.f32(float %Val)
4686 declare double @llvm.sin.f64(double %Val)
4687 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
4688 declare fp128 @llvm.sin.f128(fp128 %Val)
4689 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
4690</pre>
4691
4692<h5>Overview:</h5>
4693
4694<p>
4695The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.
4696</p>
4697
4698<h5>Arguments:</h5>
4699
4700<p>
4701The argument and return value are floating point numbers of the same type.
4702</p>
4703
4704<h5>Semantics:</h5>
4705
4706<p>
4707This function returns the sine of the specified operand, returning the
4708same values as the libm <tt>sin</tt> functions would, and handles error
Dan Gohmaneaba92e2007-10-17 18:05:13 +0000[diff] [blame]4709conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]4710</div>
4711
4712<!-- _______________________________________________________________________ -->
4713<div class="doc_subsubsection">
4714 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
4715</div>
4716
4717<div class="doc_text">
4718
4719<h5>Syntax:</h5>
4720<p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
4721floating point or vector of floating point type. Not all targets support all
4722types however.
4723<pre>
4724 declare float @llvm.cos.f32(float %Val)
4725 declare double @llvm.cos.f64(double %Val)
4726 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
4727 declare fp128 @llvm.cos.f128(fp128 %Val)
4728 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
4729</pre>
4730
4731<h5>Overview:</h5>
4732
4733<p>
4734The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.
4735</p>
4736
4737<h5>Arguments:</h5>
4738
4739<p>
4740The argument and return value are floating point numbers of the same type.
4741</p>
4742
4743<h5>Semantics:</h5>
4744
4745<p>
4746This function returns the cosine of the specified operand, returning the
4747same values as the libm <tt>cos</tt> functions would, and handles error
Dan Gohmaneaba92e2007-10-17 18:05:13 +0000[diff] [blame]4748conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]4749</div>
4750
4751<!-- _______________________________________________________________________ -->
4752<div class="doc_subsubsection">
4753 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
4754</div>
4755
4756<div class="doc_text">
4757
4758<h5>Syntax:</h5>
4759<p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
4760floating point or vector of floating point type. Not all targets support all
4761types however.
4762<pre>
4763 declare float @llvm.pow.f32(float %Val, float %Power)
4764 declare double @llvm.pow.f64(double %Val, double %Power)
4765 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
4766 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
4767 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
4768</pre>
4769
4770<h5>Overview:</h5>
4771
4772<p>
4773The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
4774specified (positive or negative) power.
4775</p>
4776
4777<h5>Arguments:</h5>
4778
4779<p>
4780The second argument is a floating point power, and the first is a value to
4781raise to that power.
4782</p>
4783
4784<h5>Semantics:</h5>
4785
4786<p>
4787This function returns the first value raised to the second power,
4788returning the
4789same values as the libm <tt>pow</tt> functions would, and handles error
Dan Gohmaneaba92e2007-10-17 18:05:13 +0000[diff] [blame]4790conditions in the same way.</p>
Dan Gohman361079c2007-10-15 20:30:11 +0000[diff] [blame]4791</div>
4792
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4793
4794<!-- ======================================================================= -->
4795<div class="doc_subsection">
4796 <a name="int_manip">Bit Manipulation Intrinsics</a>
4797</div>
4798
4799<div class="doc_text">
4800<p>
4801LLVM provides intrinsics for a few important bit manipulation operations.
4802These allow efficient code generation for some algorithms.
4803</p>
4804
4805</div>
4806
4807<!-- _______________________________________________________________________ -->
4808<div class="doc_subsubsection">
4809 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
4810</div>
4811
4812<div class="doc_text">
4813
4814<h5>Syntax:</h5>
4815<p>This is an overloaded intrinsic function. You can use bswap on any integer
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4816type that is an even number of bytes (i.e. BitWidth % 16 == 0).
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4817<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4818 declare i16 @llvm.bswap.i16(i16 &lt;id&gt;)
4819 declare i32 @llvm.bswap.i32(i32 &lt;id&gt;)
4820 declare i64 @llvm.bswap.i64(i64 &lt;id&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4821</pre>
4822
4823<h5>Overview:</h5>
4824
4825<p>
4826The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
4827values with an even number of bytes (positive multiple of 16 bits). These are
4828useful for performing operations on data that is not in the target's native
4829byte order.
4830</p>
4831
4832<h5>Semantics:</h5>
4833
4834<p>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4835The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4836and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
4837intrinsic returns an i32 value that has the four bytes of the input i32
4838swapped, 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]4839i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48</tt>,
4840<tt>llvm.bswap.i64</tt> and other intrinsics extend this concept to
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4841additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
4842</p>
4843
4844</div>
4845
4846<!-- _______________________________________________________________________ -->
4847<div class="doc_subsubsection">
4848 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
4849</div>
4850
4851<div class="doc_text">
4852
4853<h5>Syntax:</h5>
4854<p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
4855width. Not all targets support all bit widths however.
4856<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4857 declare i8 @llvm.ctpop.i8 (i8 &lt;src&gt;)
4858 declare i16 @llvm.ctpop.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4859 declare i32 @llvm.ctpop.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4860 declare i64 @llvm.ctpop.i64(i64 &lt;src&gt;)
4861 declare i256 @llvm.ctpop.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4862</pre>
4863
4864<h5>Overview:</h5>
4865
4866<p>
4867The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
4868value.
4869</p>
4870
4871<h5>Arguments:</h5>
4872
4873<p>
4874The only argument is the value to be counted. The argument may be of any
4875integer type. The return type must match the argument type.
4876</p>
4877
4878<h5>Semantics:</h5>
4879
4880<p>
4881The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
4882</p>
4883</div>
4884
4885<!-- _______________________________________________________________________ -->
4886<div class="doc_subsubsection">
4887 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
4888</div>
4889
4890<div class="doc_text">
4891
4892<h5>Syntax:</h5>
4893<p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
4894integer bit width. Not all targets support all bit widths however.
4895<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4896 declare i8 @llvm.ctlz.i8 (i8 &lt;src&gt;)
4897 declare i16 @llvm.ctlz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4898 declare i32 @llvm.ctlz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4899 declare i64 @llvm.ctlz.i64(i64 &lt;src&gt;)
4900 declare i256 @llvm.ctlz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4901</pre>
4902
4903<h5>Overview:</h5>
4904
4905<p>
4906The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
4907leading zeros in a variable.
4908</p>
4909
4910<h5>Arguments:</h5>
4911
4912<p>
4913The only argument is the value to be counted. The argument may be of any
4914integer type. The return type must match the argument type.
4915</p>
4916
4917<h5>Semantics:</h5>
4918
4919<p>
4920The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
4921in a variable. If the src == 0 then the result is the size in bits of the type
4922of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
4923</p>
4924</div>
4925
4926
4927
4928<!-- _______________________________________________________________________ -->
4929<div class="doc_subsubsection">
4930 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
4931</div>
4932
4933<div class="doc_text">
4934
4935<h5>Syntax:</h5>
4936<p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
4937integer bit width. Not all targets support all bit widths however.
4938<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4939 declare i8 @llvm.cttz.i8 (i8 &lt;src&gt;)
4940 declare i16 @llvm.cttz.i16(i16 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4941 declare i32 @llvm.cttz.i32(i32 &lt;src&gt;)
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4942 declare i64 @llvm.cttz.i64(i64 &lt;src&gt;)
4943 declare i256 @llvm.cttz.i256(i256 &lt;src&gt;)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4944</pre>
4945
4946<h5>Overview:</h5>
4947
4948<p>
4949The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
4950trailing zeros.
4951</p>
4952
4953<h5>Arguments:</h5>
4954
4955<p>
4956The only argument is the value to be counted. The argument may be of any
4957integer type. The return type must match the argument type.
4958</p>
4959
4960<h5>Semantics:</h5>
4961
4962<p>
4963The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
4964in a variable. If the src == 0 then the result is the size in bits of the type
4965of src. For example, <tt>llvm.cttz(2) = 1</tt>.
4966</p>
4967</div>
4968
4969<!-- _______________________________________________________________________ -->
4970<div class="doc_subsubsection">
4971 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
4972</div>
4973
4974<div class="doc_text">
4975
4976<h5>Syntax:</h5>
4977<p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
4978on any integer bit width.
4979<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]4980 declare i17 @llvm.part.select.i17 (i17 %val, i32 %loBit, i32 %hiBit)
4981 declare i29 @llvm.part.select.i29 (i29 %val, i32 %loBit, i32 %hiBit)
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]4982</pre>
4983
4984<h5>Overview:</h5>
4985<p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
4986range of bits from an integer value and returns them in the same bit width as
4987the original value.</p>
4988
4989<h5>Arguments:</h5>
4990<p>The first argument, <tt>%val</tt> and the result may be integer types of
4991any bit width but they must have the same bit width. The second and third
4992arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
4993
4994<h5>Semantics:</h5>
4995<p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
4996of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
4997<tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
4998operates in forward mode.</p>
4999<p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
5000right by <tt>%loBit</tt> bits and then ANDing it with a mask with
5001only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
5002<ol>
5003 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
5004 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
5005 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
5006 to determine the number of bits to retain.</li>
5007 <li>A mask of the retained bits is created by shifting a -1 value.</li>
5008 <li>The mask is ANDed with <tt>%val</tt> to produce the result.
5009</ol>
5010<p>In reverse mode, a similar computation is made except that the bits are
5011returned in the reverse order. So, for example, if <tt>X</tt> has the value
5012<tt>i16 0x0ACF (101011001111)</tt> and we apply
5013<tt>part.select(i16 X, 8, 3)</tt> to it, we get back the value
5014<tt>i16 0x0026 (000000100110)</tt>.</p>
5015</div>
5016
5017<div class="doc_subsubsection">
5018 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
5019</div>
5020
5021<div class="doc_text">
5022
5023<h5>Syntax:</h5>
5024<p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
5025on any integer bit width.
5026<pre>
Chandler Carrutha228e392007-08-04 01:51:18 +0000[diff] [blame]5027 declare i17 @llvm.part.set.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
5028 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]5029</pre>
5030
5031<h5>Overview:</h5>
5032<p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
5033of bits in an integer value with another integer value. It returns the integer
5034with the replaced bits.</p>
5035
5036<h5>Arguments:</h5>
5037<p>The first argument, <tt>%val</tt> and the result may be integer types of
5038any bit width but they must have the same bit width. <tt>%val</tt> is the value
5039whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
5040integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
5041type since they specify only a bit index.</p>
5042
5043<h5>Semantics:</h5>
5044<p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
5045of operation: forwards and reverse. If <tt>%lo</tt> is greater than
5046<tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
5047operates in forward mode.</p>
5048<p>For both modes, the <tt>%repl</tt> value is prepared for use by either
5049truncating it down to the size of the replacement area or zero extending it
5050up to that size.</p>
5051<p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
5052are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
5053in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
5054to the <tt>%hi</tt>th bit.
5055<p>In reverse mode, a similar computation is made except that the bits are
5056reversed. That is, the <tt>0</tt>th bit in <tt>%repl</tt> replaces the
5057<tt>%hi</tt> bit in <tt>%val</tt> and etc. down to the <tt>%lo</tt>th bit.
5058<h5>Examples:</h5>
5059<pre>
5060 llvm.part.set(0xFFFF, 0, 4, 7) -&gt; 0xFF0F
5061 llvm.part.set(0xFFFF, 0, 7, 4) -&gt; 0xFF0F
5062 llvm.part.set(0xFFFF, 1, 7, 4) -&gt; 0xFF8F
5063 llvm.part.set(0xFFFF, F, 8, 3) -&gt; 0xFFE7
5064 llvm.part.set(0xFFFF, 0, 3, 8) -&gt; 0xFE07
5065</pre>
5066</div>
5067
5068<!-- ======================================================================= -->
5069<div class="doc_subsection">
5070 <a name="int_debugger">Debugger Intrinsics</a>
5071</div>
5072
5073<div class="doc_text">
5074<p>
5075The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
5076are described in the <a
5077href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
5078Debugging</a> document.
5079</p>
5080</div>
5081
5082
5083<!-- ======================================================================= -->
5084<div class="doc_subsection">
5085 <a name="int_eh">Exception Handling Intrinsics</a>
5086</div>
5087
5088<div class="doc_text">
5089<p> The LLVM exception handling intrinsics (which all start with
5090<tt>llvm.eh.</tt> prefix), are described in the <a
5091href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
5092Handling</a> document. </p>
5093</div>
5094
5095<!-- ======================================================================= -->
5096<div class="doc_subsection">
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5097 <a name="int_trampoline">Trampoline Intrinsic</a>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5098</div>
5099
5100<div class="doc_text">
5101<p>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5102 This intrinsic makes it possible to excise one parameter, marked with
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5103 the <tt>nest</tt> attribute, from a function. The result is a callable
5104 function pointer lacking the nest parameter - the caller does not need
5105 to provide a value for it. Instead, the value to use is stored in
5106 advance in a "trampoline", a block of memory usually allocated
5107 on the stack, which also contains code to splice the nest value into the
5108 argument list. This is used to implement the GCC nested function address
5109 extension.
5110</p>
5111<p>
5112 For example, if the function is
5113 <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]5114 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]5115<pre>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5116 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
5117 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
5118 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
5119 %fp = bitcast i8* %p to i32 (i32, i32)*
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5120</pre>
Bill Wendlinge262b4c2007-09-22 09:23:55 +0000[diff] [blame]5121 <p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
5122 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5123</div>
5124
5125<!-- _______________________________________________________________________ -->
5126<div class="doc_subsubsection">
5127 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
5128</div>
5129<div class="doc_text">
5130<h5>Syntax:</h5>
5131<pre>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5132declare 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]5133</pre>
5134<h5>Overview:</h5>
5135<p>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5136 This fills the memory pointed to by <tt>tramp</tt> with code
5137 and returns a function pointer suitable for executing it.
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5138</p>
5139<h5>Arguments:</h5>
5140<p>
5141 The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
5142 pointers. The <tt>tramp</tt> argument must point to a sufficiently large
5143 and sufficiently aligned block of memory; this memory is written to by the
Duncan Sands35012212007-08-22 23:39:54 +0000[diff] [blame]5144 intrinsic. Note that the size and the alignment are target-specific - LLVM
5145 currently provides no portable way of determining them, so a front-end that
5146 generates this intrinsic needs to have some target-specific knowledge.
5147 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]5148</p>
5149<h5>Semantics:</h5>
5150<p>
5151 The block of memory pointed to by <tt>tramp</tt> is filled with target
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5152 dependent code, turning it into a function. A pointer to this function is
5153 returned, but needs to be bitcast to an
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5154 <a href="#int_trampoline">appropriate function pointer type</a>
Duncan Sands7407a9f2007-09-11 14:10:23 +0000[diff] [blame]5155 before being called. The new function's signature is the same as that of
5156 <tt>func</tt> with any arguments marked with the <tt>nest</tt> attribute
5157 removed. At most one such <tt>nest</tt> argument is allowed, and it must be
5158 of pointer type. Calling the new function is equivalent to calling
5159 <tt>func</tt> with the same argument list, but with <tt>nval</tt> used for the
5160 missing <tt>nest</tt> argument. If, after calling
5161 <tt>llvm.init.trampoline</tt>, the memory pointed to by <tt>tramp</tt> is
5162 modified, then the effect of any later call to the returned function pointer is
5163 undefined.
Duncan Sands38947cd2007-07-27 12:58:54 +0000[diff] [blame]5164</p>
5165</div>
5166
5167<!-- ======================================================================= -->
5168<div class="doc_subsection">
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5169 <a name="int_general">General Intrinsics</a>
5170</div>
5171
5172<div class="doc_text">
5173<p> This class of intrinsics is designed to be generic and has
5174no specific purpose. </p>
5175</div>
5176
5177<!-- _______________________________________________________________________ -->
5178<div class="doc_subsubsection">
5179 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
5180</div>
5181
5182<div class="doc_text">
5183
5184<h5>Syntax:</h5>
5185<pre>
5186 declare void @llvm.var.annotation(i8* &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5187</pre>
5188
5189<h5>Overview:</h5>
5190
5191<p>
5192The '<tt>llvm.var.annotation</tt>' intrinsic
5193</p>
5194
5195<h5>Arguments:</h5>
5196
5197<p>
5198The first argument is a pointer to a value, the second is a pointer to a
5199global string, the third is a pointer to a global string which is the source
5200file name, and the last argument is the line number.
5201</p>
5202
5203<h5>Semantics:</h5>
5204
5205<p>
5206This intrinsic allows annotation of local variables with arbitrary strings.
5207This can be useful for special purpose optimizations that want to look for these
5208 annotations. These have no other defined use, they are ignored by code
5209 generation and optimization.
5210</div>
5211
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]5212<!-- _______________________________________________________________________ -->
5213<div class="doc_subsubsection">
Tanya Lattnerc9869b12007-09-21 23:57:59 +0000[diff] [blame]5214 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]5215</div>
5216
5217<div class="doc_text">
5218
5219<h5>Syntax:</h5>
Tanya Lattnere545be72007-09-21 23:56:27 +0000[diff] [blame]5220<p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
5221any integer bit width.
5222</p>
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]5223<pre>
Tanya Lattner09161fe2007-09-22 00:03:01 +0000[diff] [blame]5224 declare i8 @llvm.annotation.i8(i8 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5225 declare i16 @llvm.annotation.i16(i16 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5226 declare i32 @llvm.annotation.i32(i32 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5227 declare i64 @llvm.annotation.i64(i64 &lt;val&gt;, i8* &lt;str&gt;, i8* &lt;str&gt;, i32 &lt;int&gt; )
5228 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]5229</pre>
5230
5231<h5>Overview:</h5>
Tanya Lattnere545be72007-09-21 23:56:27 +0000[diff] [blame]5232
5233<p>
5234The '<tt>llvm.annotation</tt>' intrinsic.
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]5235</p>
5236
5237<h5>Arguments:</h5>
5238
5239<p>
5240The first argument is an integer value (result of some expression),
5241the second is a pointer to a global string, the third is a pointer to a global
5242string which is the source file name, and the last argument is the line number.
Tanya Lattnere545be72007-09-21 23:56:27 +0000[diff] [blame]5243It returns the value of the first argument.
Tanya Lattnerb306a9e2007-09-21 22:59:12 +0000[diff] [blame]5244</p>
5245
5246<h5>Semantics:</h5>
5247
5248<p>
5249This intrinsic allows annotations to be put on arbitrary expressions
5250with arbitrary strings. This can be useful for special purpose optimizations
5251that want to look for these annotations. These have no other defined use, they
5252are ignored by code generation and optimization.
5253</div>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5254
5255<!-- *********************************************************************** -->
5256<hr>
5257<address>
5258 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
5259 src="http://jigsaw.w3.org/css-validator/images/vcss" alt="Valid CSS!"></a>
5260 <a href="http://validator.w3.org/check/referer"><img
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]5261 src="http://www.w3.org/Icons/valid-html401" alt="Valid HTML 4.01!"></a>
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5262
5263 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
5264 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
5265 Last modified: $Date$
5266</address>
Chris Lattner08497ce2008-01-04 04:33:49 +0000[diff] [blame]5267
Dan Gohmanf17a25c2007-07-18 16:29:46 +0000[diff] [blame]5268</body>
5269</html>