1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN"
2 "http://www.w3.org/TR/html4/strict.dtd">
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">
15 <div class="doc_title"> LLVM Language Reference Manual </div>
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>
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="#namedtypes">Named Types</a></li>
26 <li><a href="#globalvars">Global Variables</a></li>
27 <li><a href="#functionstructure">Functions</a></li>
28 <li><a href="#aliasstructure">Aliases</a></li>
29 <li><a href="#paramattrs">Parameter Attributes</a></li>
30 <li><a href="#fnattrs">Function Attributes</a></li>
31 <li><a href="#gc">Garbage Collector Names</a></li>
32 <li><a href="#moduleasm">Module-Level Inline Assembly</a></li>
33 <li><a href="#datalayout">Data Layout</a></li>
36 <li><a href="#typesystem">Type System</a>
38 <li><a href="#t_classifications">Type Classifications</a></li>
39 <li><a href="#t_primitive">Primitive Types</a>
41 <li><a href="#t_floating">Floating Point Types</a></li>
42 <li><a href="#t_void">Void Type</a></li>
43 <li><a href="#t_label">Label Type</a></li>
46 <li><a href="#t_derived">Derived Types</a>
48 <li><a href="#t_integer">Integer Type</a></li>
49 <li><a href="#t_array">Array Type</a></li>
50 <li><a href="#t_function">Function Type</a></li>
51 <li><a href="#t_pointer">Pointer Type</a></li>
52 <li><a href="#t_struct">Structure Type</a></li>
53 <li><a href="#t_pstruct">Packed Structure Type</a></li>
54 <li><a href="#t_vector">Vector Type</a></li>
55 <li><a href="#t_opaque">Opaque Type</a></li>
58 <li><a href="#t_uprefs">Type Up-references</a></li>
61 <li><a href="#constants">Constants</a>
63 <li><a href="#simpleconstants">Simple Constants</a></li>
64 <li><a href="#aggregateconstants">Aggregate Constants</a></li>
65 <li><a href="#globalconstants">Global Variable and Function Addresses</a></li>
66 <li><a href="#undefvalues">Undefined Values</a></li>
67 <li><a href="#constantexprs">Constant Expressions</a></li>
70 <li><a href="#othervalues">Other Values</a>
72 <li><a href="#inlineasm">Inline Assembler Expressions</a></li>
75 <li><a href="#instref">Instruction Reference</a>
77 <li><a href="#terminators">Terminator Instructions</a>
79 <li><a href="#i_ret">'<tt>ret</tt>' Instruction</a></li>
80 <li><a href="#i_br">'<tt>br</tt>' Instruction</a></li>
81 <li><a href="#i_switch">'<tt>switch</tt>' Instruction</a></li>
82 <li><a href="#i_invoke">'<tt>invoke</tt>' Instruction</a></li>
83 <li><a href="#i_unwind">'<tt>unwind</tt>' Instruction</a></li>
84 <li><a href="#i_unreachable">'<tt>unreachable</tt>' Instruction</a></li>
87 <li><a href="#binaryops">Binary Operations</a>
89 <li><a href="#i_add">'<tt>add</tt>' Instruction</a></li>
90 <li><a href="#i_sub">'<tt>sub</tt>' Instruction</a></li>
91 <li><a href="#i_mul">'<tt>mul</tt>' Instruction</a></li>
92 <li><a href="#i_udiv">'<tt>udiv</tt>' Instruction</a></li>
93 <li><a href="#i_sdiv">'<tt>sdiv</tt>' Instruction</a></li>
94 <li><a href="#i_fdiv">'<tt>fdiv</tt>' Instruction</a></li>
95 <li><a href="#i_urem">'<tt>urem</tt>' Instruction</a></li>
96 <li><a href="#i_srem">'<tt>srem</tt>' Instruction</a></li>
97 <li><a href="#i_frem">'<tt>frem</tt>' Instruction</a></li>
100 <li><a href="#bitwiseops">Bitwise Binary Operations</a>
102 <li><a href="#i_shl">'<tt>shl</tt>' Instruction</a></li>
103 <li><a href="#i_lshr">'<tt>lshr</tt>' Instruction</a></li>
104 <li><a href="#i_ashr">'<tt>ashr</tt>' Instruction</a></li>
105 <li><a href="#i_and">'<tt>and</tt>' Instruction</a></li>
106 <li><a href="#i_or">'<tt>or</tt>' Instruction</a></li>
107 <li><a href="#i_xor">'<tt>xor</tt>' Instruction</a></li>
110 <li><a href="#vectorops">Vector Operations</a>
112 <li><a href="#i_extractelement">'<tt>extractelement</tt>' Instruction</a></li>
113 <li><a href="#i_insertelement">'<tt>insertelement</tt>' Instruction</a></li>
114 <li><a href="#i_shufflevector">'<tt>shufflevector</tt>' Instruction</a></li>
117 <li><a href="#aggregateops">Aggregate Operations</a>
119 <li><a href="#i_extractvalue">'<tt>extractvalue</tt>' Instruction</a></li>
120 <li><a href="#i_insertvalue">'<tt>insertvalue</tt>' Instruction</a></li>
123 <li><a href="#memoryops">Memory Access and Addressing Operations</a>
125 <li><a href="#i_malloc">'<tt>malloc</tt>' Instruction</a></li>
126 <li><a href="#i_free">'<tt>free</tt>' Instruction</a></li>
127 <li><a href="#i_alloca">'<tt>alloca</tt>' Instruction</a></li>
128 <li><a href="#i_load">'<tt>load</tt>' Instruction</a></li>
129 <li><a href="#i_store">'<tt>store</tt>' Instruction</a></li>
130 <li><a href="#i_getelementptr">'<tt>getelementptr</tt>' Instruction</a></li>
133 <li><a href="#convertops">Conversion Operations</a>
135 <li><a href="#i_trunc">'<tt>trunc .. to</tt>' Instruction</a></li>
136 <li><a href="#i_zext">'<tt>zext .. to</tt>' Instruction</a></li>
137 <li><a href="#i_sext">'<tt>sext .. to</tt>' Instruction</a></li>
138 <li><a href="#i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a></li>
139 <li><a href="#i_fpext">'<tt>fpext .. to</tt>' Instruction</a></li>
140 <li><a href="#i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a></li>
141 <li><a href="#i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a></li>
142 <li><a href="#i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a></li>
143 <li><a href="#i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a></li>
144 <li><a href="#i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a></li>
145 <li><a href="#i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a></li>
146 <li><a href="#i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a></li>
149 <li><a href="#otherops">Other Operations</a>
151 <li><a href="#i_icmp">'<tt>icmp</tt>' Instruction</a></li>
152 <li><a href="#i_fcmp">'<tt>fcmp</tt>' Instruction</a></li>
153 <li><a href="#i_vicmp">'<tt>vicmp</tt>' Instruction</a></li>
154 <li><a href="#i_vfcmp">'<tt>vfcmp</tt>' Instruction</a></li>
155 <li><a href="#i_phi">'<tt>phi</tt>' Instruction</a></li>
156 <li><a href="#i_select">'<tt>select</tt>' Instruction</a></li>
157 <li><a href="#i_call">'<tt>call</tt>' Instruction</a></li>
158 <li><a href="#i_va_arg">'<tt>va_arg</tt>' Instruction</a></li>
163 <li><a href="#intrinsics">Intrinsic Functions</a>
165 <li><a href="#int_varargs">Variable Argument Handling Intrinsics</a>
167 <li><a href="#int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a></li>
168 <li><a href="#int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a></li>
169 <li><a href="#int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a></li>
172 <li><a href="#int_gc">Accurate Garbage Collection Intrinsics</a>
174 <li><a href="#int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a></li>
175 <li><a href="#int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a></li>
176 <li><a href="#int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a></li>
179 <li><a href="#int_codegen">Code Generator Intrinsics</a>
181 <li><a href="#int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a></li>
182 <li><a href="#int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a></li>
183 <li><a href="#int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a></li>
184 <li><a href="#int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a></li>
185 <li><a href="#int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a></li>
186 <li><a href="#int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a></li>
187 <li><a href="#int_readcyclecounter"><tt>llvm.readcyclecounter</tt>' Intrinsic</a></li>
190 <li><a href="#int_libc">Standard C Library Intrinsics</a>
192 <li><a href="#int_memcpy">'<tt>llvm.memcpy.*</tt>' Intrinsic</a></li>
193 <li><a href="#int_memmove">'<tt>llvm.memmove.*</tt>' Intrinsic</a></li>
194 <li><a href="#int_memset">'<tt>llvm.memset.*</tt>' Intrinsic</a></li>
195 <li><a href="#int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a></li>
196 <li><a href="#int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a></li>
197 <li><a href="#int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a></li>
198 <li><a href="#int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a></li>
199 <li><a href="#int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a></li>
202 <li><a href="#int_manip">Bit Manipulation Intrinsics</a>
204 <li><a href="#int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a></li>
205 <li><a href="#int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic </a></li>
206 <li><a href="#int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic </a></li>
207 <li><a href="#int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic </a></li>
208 <li><a href="#int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic </a></li>
209 <li><a href="#int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic </a></li>
212 <li><a href="#int_overflow">Arithmetic with Overflow Intrinsics</a>
214 <li><a href="#int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt> Intrinsics</a></li>
215 <li><a href="#int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt> Intrinsics</a></li>
216 <li><a href="#int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt> Intrinsics</a></li>
217 <li><a href="#int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt> Intrinsics</a></li>
218 <li><a href="#int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt> Intrinsics</a></li>
221 <li><a href="#int_debugger">Debugger intrinsics</a></li>
222 <li><a href="#int_eh">Exception Handling intrinsics</a></li>
223 <li><a href="#int_trampoline">Trampoline Intrinsic</a>
225 <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
228 <li><a href="#int_atomics">Atomic intrinsics</a>
230 <li><a href="#int_memory_barrier"><tt>llvm.memory_barrier</tt></a></li>
231 <li><a href="#int_atomic_cmp_swap"><tt>llvm.atomic.cmp.swap</tt></a></li>
232 <li><a href="#int_atomic_swap"><tt>llvm.atomic.swap</tt></a></li>
233 <li><a href="#int_atomic_load_add"><tt>llvm.atomic.load.add</tt></a></li>
234 <li><a href="#int_atomic_load_sub"><tt>llvm.atomic.load.sub</tt></a></li>
235 <li><a href="#int_atomic_load_and"><tt>llvm.atomic.load.and</tt></a></li>
236 <li><a href="#int_atomic_load_nand"><tt>llvm.atomic.load.nand</tt></a></li>
237 <li><a href="#int_atomic_load_or"><tt>llvm.atomic.load.or</tt></a></li>
238 <li><a href="#int_atomic_load_xor"><tt>llvm.atomic.load.xor</tt></a></li>
239 <li><a href="#int_atomic_load_max"><tt>llvm.atomic.load.max</tt></a></li>
240 <li><a href="#int_atomic_load_min"><tt>llvm.atomic.load.min</tt></a></li>
241 <li><a href="#int_atomic_load_umax"><tt>llvm.atomic.load.umax</tt></a></li>
242 <li><a href="#int_atomic_load_umin"><tt>llvm.atomic.load.umin</tt></a></li>
245 <li><a href="#int_general">General intrinsics</a>
247 <li><a href="#int_var_annotation">
248 '<tt>llvm.var.annotation</tt>' Intrinsic</a></li>
249 <li><a href="#int_annotation">
250 '<tt>llvm.annotation.*</tt>' Intrinsic</a></li>
251 <li><a href="#int_trap">
252 '<tt>llvm.trap</tt>' Intrinsic</a></li>
253 <li><a href="#int_stackprotector">
254 '<tt>llvm.stackprotector</tt>' Intrinsic</a></li>
261 <div class="doc_author">
262 <p>Written by <a href="mailto:sabre@nondot.org">Chris Lattner</a>
263 and <a href="mailto:vadve@cs.uiuc.edu">Vikram Adve</a></p>
266 <!-- *********************************************************************** -->
267 <div class="doc_section"> <a name="abstract">Abstract </a></div>
268 <!-- *********************************************************************** -->
270 <div class="doc_text">
271 <p>This document is a reference manual for the LLVM assembly language.
272 LLVM is a Static Single Assignment (SSA) based representation that provides
273 type safety, low-level operations, flexibility, and the capability of
274 representing 'all' high-level languages cleanly. It is the common code
275 representation used throughout all phases of the LLVM compilation
279 <!-- *********************************************************************** -->
280 <div class="doc_section"> <a name="introduction">Introduction</a> </div>
281 <!-- *********************************************************************** -->
283 <div class="doc_text">
285 <p>The LLVM code representation is designed to be used in three
286 different forms: as an in-memory compiler IR, as an on-disk bitcode
287 representation (suitable for fast loading by a Just-In-Time compiler),
288 and as a human readable assembly language representation. This allows
289 LLVM to provide a powerful intermediate representation for efficient
290 compiler transformations and analysis, while providing a natural means
291 to debug and visualize the transformations. The three different forms
292 of LLVM are all equivalent. This document describes the human readable
293 representation and notation.</p>
295 <p>The LLVM representation aims to be light-weight and low-level
296 while being expressive, typed, and extensible at the same time. It
297 aims to be a "universal IR" of sorts, by being at a low enough level
298 that high-level ideas may be cleanly mapped to it (similar to how
299 microprocessors are "universal IR's", allowing many source languages to
300 be mapped to them). By providing type information, LLVM can be used as
301 the target of optimizations: for example, through pointer analysis, it
302 can be proven that a C automatic variable is never accessed outside of
303 the current function... allowing it to be promoted to a simple SSA
304 value instead of a memory location.</p>
308 <!-- _______________________________________________________________________ -->
309 <div class="doc_subsubsection"> <a name="wellformed">Well-Formedness</a> </div>
311 <div class="doc_text">
313 <p>It is important to note that this document describes 'well formed'
314 LLVM assembly language. There is a difference between what the parser
315 accepts and what is considered 'well formed'. For example, the
316 following instruction is syntactically okay, but not well formed:</p>
318 <div class="doc_code">
320 %x = <a href="#i_add">add</a> i32 1, %x
324 <p>...because the definition of <tt>%x</tt> does not dominate all of
325 its uses. The LLVM infrastructure provides a verification pass that may
326 be used to verify that an LLVM module is well formed. This pass is
327 automatically run by the parser after parsing input assembly and by
328 the optimizer before it outputs bitcode. The violations pointed out
329 by the verifier pass indicate bugs in transformation passes or input to
333 <!-- Describe the typesetting conventions here. -->
335 <!-- *********************************************************************** -->
336 <div class="doc_section"> <a name="identifiers">Identifiers</a> </div>
337 <!-- *********************************************************************** -->
339 <div class="doc_text">
341 <p>LLVM identifiers come in two basic types: global and local. Global
342 identifiers (functions, global variables) begin with the @ character. Local
343 identifiers (register names, types) begin with the % character. Additionally,
344 there are three different formats for identifiers, for different purposes:</p>
347 <li>Named values are represented as a string of characters with their prefix.
348 For example, %foo, @DivisionByZero, %a.really.long.identifier. The actual
349 regular expression used is '<tt>[%@][a-zA-Z$._][a-zA-Z$._0-9]*</tt>'.
350 Identifiers which require other characters in their names can be surrounded
351 with quotes. Special characters may be escaped using "\xx" where xx is the
352 ASCII code for the character in hexadecimal. In this way, any character can
353 be used in a name value, even quotes themselves.
355 <li>Unnamed values are represented as an unsigned numeric value with their
356 prefix. For example, %12, @2, %44.</li>
358 <li>Constants, which are described in a <a href="#constants">section about
359 constants</a>, below.</li>
362 <p>LLVM requires that values start with a prefix for two reasons: Compilers
363 don't need to worry about name clashes with reserved words, and the set of
364 reserved words may be expanded in the future without penalty. Additionally,
365 unnamed identifiers allow a compiler to quickly come up with a temporary
366 variable without having to avoid symbol table conflicts.</p>
368 <p>Reserved words in LLVM are very similar to reserved words in other
369 languages. There are keywords for different opcodes
370 ('<tt><a href="#i_add">add</a></tt>',
371 '<tt><a href="#i_bitcast">bitcast</a></tt>',
372 '<tt><a href="#i_ret">ret</a></tt>', etc...), for primitive type names ('<tt><a
373 href="#t_void">void</a></tt>', '<tt><a href="#t_primitive">i32</a></tt>', etc...),
374 and others. These reserved words cannot conflict with variable names, because
375 none of them start with a prefix character ('%' or '@').</p>
377 <p>Here is an example of LLVM code to multiply the integer variable
378 '<tt>%X</tt>' by 8:</p>
382 <div class="doc_code">
384 %result = <a href="#i_mul">mul</a> i32 %X, 8
388 <p>After strength reduction:</p>
390 <div class="doc_code">
392 %result = <a href="#i_shl">shl</a> i32 %X, i8 3
396 <p>And the hard way:</p>
398 <div class="doc_code">
400 <a href="#i_add">add</a> i32 %X, %X <i>; yields {i32}:%0</i>
401 <a href="#i_add">add</a> i32 %0, %0 <i>; yields {i32}:%1</i>
402 %result = <a href="#i_add">add</a> i32 %1, %1
406 <p>This last way of multiplying <tt>%X</tt> by 8 illustrates several
407 important lexical features of LLVM:</p>
411 <li>Comments are delimited with a '<tt>;</tt>' and go until the end of
414 <li>Unnamed temporaries are created when the result of a computation is not
415 assigned to a named value.</li>
417 <li>Unnamed temporaries are numbered sequentially</li>
421 <p>...and it also shows a convention that we follow in this document. When
422 demonstrating instructions, we will follow an instruction with a comment that
423 defines the type and name of value produced. Comments are shown in italic
428 <!-- *********************************************************************** -->
429 <div class="doc_section"> <a name="highlevel">High Level Structure</a> </div>
430 <!-- *********************************************************************** -->
432 <!-- ======================================================================= -->
433 <div class="doc_subsection"> <a name="modulestructure">Module Structure</a>
436 <div class="doc_text">
438 <p>LLVM programs are composed of "Module"s, each of which is a
439 translation unit of the input programs. Each module consists of
440 functions, global variables, and symbol table entries. Modules may be
441 combined together with the LLVM linker, which merges function (and
442 global variable) definitions, resolves forward declarations, and merges
443 symbol table entries. Here is an example of the "hello world" module:</p>
445 <div class="doc_code">
446 <pre><i>; Declare the string constant as a global constant...</i>
447 <a href="#identifiers">@.LC0</a> = <a href="#linkage_internal">internal</a> <a
448 href="#globalvars">constant</a> <a href="#t_array">[13 x i8]</a> c"hello world\0A\00" <i>; [13 x i8]*</i>
450 <i>; External declaration of the puts function</i>
451 <a href="#functionstructure">declare</a> i32 @puts(i8 *) <i>; i32(i8 *)* </i>
453 <i>; Definition of main function</i>
454 define i32 @main() { <i>; i32()* </i>
455 <i>; Convert [13 x i8]* to i8 *...</i>
457 href="#i_getelementptr">getelementptr</a> [13 x i8]* @.LC0, i64 0, i64 0 <i>; i8 *</i>
459 <i>; Call puts function to write out the string to stdout...</i>
461 href="#i_call">call</a> i32 @puts(i8 * %cast210) <i>; i32</i>
463 href="#i_ret">ret</a> i32 0<br>}<br>
467 <p>This example is made up of a <a href="#globalvars">global variable</a>
468 named "<tt>.LC0</tt>", an external declaration of the "<tt>puts</tt>"
469 function, and a <a href="#functionstructure">function definition</a>
470 for "<tt>main</tt>".</p>
472 <p>In general, a module is made up of a list of global values,
473 where both functions and global variables are global values. Global values are
474 represented by a pointer to a memory location (in this case, a pointer to an
475 array of char, and a pointer to a function), and have one of the following <a
476 href="#linkage">linkage types</a>.</p>
480 <!-- ======================================================================= -->
481 <div class="doc_subsection">
482 <a name="linkage">Linkage Types</a>
485 <div class="doc_text">
488 All Global Variables and Functions have one of the following types of linkage:
493 <dt><tt><b><a name="linkage_private">private</a></b></tt>: </dt>
495 <dd>Global values with private linkage are only directly accessible by
496 objects in the current module. In particular, linking code into a module with
497 an private global value may cause the private to be renamed as necessary to
498 avoid collisions. Because the symbol is private to the module, all
499 references can be updated. This doesn't show up in any symbol table in the
503 <dt><tt><b><a name="linkage_internal">internal</a></b></tt>: </dt>
505 <dd> Similar to private, but the value shows as a local symbol (STB_LOCAL in
506 the case of ELF) in the object file. This corresponds to the notion of the
507 '<tt>static</tt>' keyword in C.
510 <dt><tt><b><a name="linkage_linkonce">linkonce</a></b></tt>: </dt>
512 <dd>Globals with "<tt>linkonce</tt>" linkage are merged with other globals of
513 the same name when linkage occurs. This is typically used to implement
514 inline functions, templates, or other code which must be generated in each
515 translation unit that uses it. Unreferenced <tt>linkonce</tt> globals are
516 allowed to be discarded.
519 <dt><tt><b><a name="linkage_common">common</a></b></tt>: </dt>
521 <dd>"<tt>common</tt>" linkage is exactly the same as <tt>linkonce</tt>
522 linkage, except that unreferenced <tt>common</tt> globals may not be
523 discarded. This is used for globals that may be emitted in multiple
524 translation units, but that are not guaranteed to be emitted into every
525 translation unit that uses them. One example of this is tentative
526 definitions in C, such as "<tt>int X;</tt>" at global scope.
529 <dt><tt><b><a name="linkage_weak">weak</a></b></tt>: </dt>
531 <dd>"<tt>weak</tt>" linkage is the same as <tt>common</tt> linkage, except
532 that some targets may choose to emit different assembly sequences for them
533 for target-dependent reasons. This is used for globals that are declared
534 "weak" in C source code.
537 <dt><tt><b><a name="linkage_appending">appending</a></b></tt>: </dt>
539 <dd>"<tt>appending</tt>" linkage may only be applied to global variables of
540 pointer to array type. When two global variables with appending linkage are
541 linked together, the two global arrays are appended together. This is the
542 LLVM, typesafe, equivalent of having the system linker append together
543 "sections" with identical names when .o files are linked.
546 <dt><tt><b><a name="linkage_externweak">extern_weak</a></b></tt>: </dt>
547 <dd>The semantics of this linkage follow the ELF object file model: the
548 symbol is weak until linked, if not linked, the symbol becomes null instead
549 of being an undefined reference.
552 <dt><tt><b><a name="linkage_external">externally visible</a></b></tt>:</dt>
554 <dd>If none of the above identifiers are used, the global is externally
555 visible, meaning that it participates in linkage and can be used to resolve
556 external symbol references.
561 The next two types of linkage are targeted for Microsoft Windows platform
562 only. They are designed to support importing (exporting) symbols from (to)
563 DLLs (Dynamic Link Libraries).
567 <dt><tt><b><a name="linkage_dllimport">dllimport</a></b></tt>: </dt>
569 <dd>"<tt>dllimport</tt>" linkage causes the compiler to reference a function
570 or variable via a global pointer to a pointer that is set up by the DLL
571 exporting the symbol. On Microsoft Windows targets, the pointer name is
572 formed by combining <code>__imp_</code> and the function or variable name.
575 <dt><tt><b><a name="linkage_dllexport">dllexport</a></b></tt>: </dt>
577 <dd>"<tt>dllexport</tt>" linkage causes the compiler to provide a global
578 pointer to a pointer in a DLL, so that it can be referenced with the
579 <tt>dllimport</tt> attribute. On Microsoft Windows targets, the pointer
580 name is formed by combining <code>__imp_</code> and the function or variable
586 <p>For example, since the "<tt>.LC0</tt>"
587 variable is defined to be internal, if another module defined a "<tt>.LC0</tt>"
588 variable and was linked with this one, one of the two would be renamed,
589 preventing a collision. Since "<tt>main</tt>" and "<tt>puts</tt>" are
590 external (i.e., lacking any linkage declarations), they are accessible
591 outside of the current module.</p>
592 <p>It is illegal for a function <i>declaration</i>
593 to have any linkage type other than "externally visible", <tt>dllimport</tt>,
594 or <tt>extern_weak</tt>.</p>
595 <p>Aliases can have only <tt>external</tt>, <tt>internal</tt> and <tt>weak</tt>
599 <!-- ======================================================================= -->
600 <div class="doc_subsection">
601 <a name="callingconv">Calling Conventions</a>
604 <div class="doc_text">
606 <p>LLVM <a href="#functionstructure">functions</a>, <a href="#i_call">calls</a>
607 and <a href="#i_invoke">invokes</a> can all have an optional calling convention
608 specified for the call. The calling convention of any pair of dynamic
609 caller/callee must match, or the behavior of the program is undefined. The
610 following calling conventions are supported by LLVM, and more may be added in
614 <dt><b>"<tt>ccc</tt>" - The C calling convention</b>:</dt>
616 <dd>This calling convention (the default if no other calling convention is
617 specified) matches the target C calling conventions. This calling convention
618 supports varargs function calls and tolerates some mismatch in the declared
619 prototype and implemented declaration of the function (as does normal C).
622 <dt><b>"<tt>fastcc</tt>" - The fast calling convention</b>:</dt>
624 <dd>This calling convention attempts to make calls as fast as possible
625 (e.g. by passing things in registers). This calling convention allows the
626 target to use whatever tricks it wants to produce fast code for the target,
627 without having to conform to an externally specified ABI (Application Binary
628 Interface). Implementations of this convention should allow arbitrary
629 <a href="CodeGenerator.html#tailcallopt">tail call optimization</a> to be
630 supported. This calling convention does not support varargs and requires the
631 prototype of all callees to exactly match the prototype of the function
635 <dt><b>"<tt>coldcc</tt>" - The cold calling convention</b>:</dt>
637 <dd>This calling convention attempts to make code in the caller as efficient
638 as possible under the assumption that the call is not commonly executed. As
639 such, these calls often preserve all registers so that the call does not break
640 any live ranges in the caller side. This calling convention does not support
641 varargs and requires the prototype of all callees to exactly match the
642 prototype of the function definition.
645 <dt><b>"<tt>cc <<em>n</em>></tt>" - Numbered convention</b>:</dt>
647 <dd>Any calling convention may be specified by number, allowing
648 target-specific calling conventions to be used. Target specific calling
649 conventions start at 64.
653 <p>More calling conventions can be added/defined on an as-needed basis, to
654 support pascal conventions or any other well-known target-independent
659 <!-- ======================================================================= -->
660 <div class="doc_subsection">
661 <a name="visibility">Visibility Styles</a>
664 <div class="doc_text">
667 All Global Variables and Functions have one of the following visibility styles:
671 <dt><b>"<tt>default</tt>" - Default style</b>:</dt>
673 <dd>On targets that use the ELF object file format, default visibility means
674 that the declaration is visible to other
675 modules and, in shared libraries, means that the declared entity may be
676 overridden. On Darwin, default visibility means that the declaration is
677 visible to other modules. Default visibility corresponds to "external
678 linkage" in the language.
681 <dt><b>"<tt>hidden</tt>" - Hidden style</b>:</dt>
683 <dd>Two declarations of an object with hidden visibility refer to the same
684 object if they are in the same shared object. Usually, hidden visibility
685 indicates that the symbol will not be placed into the dynamic symbol table,
686 so no other module (executable or shared library) can reference it
690 <dt><b>"<tt>protected</tt>" - Protected style</b>:</dt>
692 <dd>On ELF, protected visibility indicates that the symbol will be placed in
693 the dynamic symbol table, but that references within the defining module will
694 bind to the local symbol. That is, the symbol cannot be overridden by another
701 <!-- ======================================================================= -->
702 <div class="doc_subsection">
703 <a name="namedtypes">Named Types</a>
706 <div class="doc_text">
708 <p>LLVM IR allows you to specify name aliases for certain types. This can make
709 it easier to read the IR and make the IR more condensed (particularly when
710 recursive types are involved). An example of a name specification is:
713 <div class="doc_code">
715 %mytype = type { %mytype*, i32 }
719 <p>You may give a name to any <a href="#typesystem">type</a> except "<a
720 href="t_void">void</a>". Type name aliases may be used anywhere a type is
721 expected with the syntax "%mytype".</p>
723 <p>Note that type names are aliases for the structural type that they indicate,
724 and that you can therefore specify multiple names for the same type. This often
725 leads to confusing behavior when dumping out a .ll file. Since LLVM IR uses
726 structural typing, the name is not part of the type. When printing out LLVM IR,
727 the printer will pick <em>one name</em> to render all types of a particular
728 shape. This means that if you have code where two different source types end up
729 having the same LLVM type, that the dumper will sometimes print the "wrong" or
730 unexpected type. This is an important design point and isn't going to
735 <!-- ======================================================================= -->
736 <div class="doc_subsection">
737 <a name="globalvars">Global Variables</a>
740 <div class="doc_text">
742 <p>Global variables define regions of memory allocated at compilation time
743 instead of run-time. Global variables may optionally be initialized, may have
744 an explicit section to be placed in, and may have an optional explicit alignment
745 specified. A variable may be defined as "thread_local", which means that it
746 will not be shared by threads (each thread will have a separated copy of the
747 variable). A variable may be defined as a global "constant," which indicates
748 that the contents of the variable will <b>never</b> be modified (enabling better
749 optimization, allowing the global data to be placed in the read-only section of
750 an executable, etc). Note that variables that need runtime initialization
751 cannot be marked "constant" as there is a store to the variable.</p>
754 LLVM explicitly allows <em>declarations</em> of global variables to be marked
755 constant, even if the final definition of the global is not. This capability
756 can be used to enable slightly better optimization of the program, but requires
757 the language definition to guarantee that optimizations based on the
758 'constantness' are valid for the translation units that do not include the
762 <p>As SSA values, global variables define pointer values that are in
763 scope (i.e. they dominate) all basic blocks in the program. Global
764 variables always define a pointer to their "content" type because they
765 describe a region of memory, and all memory objects in LLVM are
766 accessed through pointers.</p>
768 <p>A global variable may be declared to reside in a target-specifc numbered
769 address space. For targets that support them, address spaces may affect how
770 optimizations are performed and/or what target instructions are used to access
771 the variable. The default address space is zero. The address space qualifier
772 must precede any other attributes.</p>
774 <p>LLVM allows an explicit section to be specified for globals. If the target
775 supports it, it will emit globals to the section specified.</p>
777 <p>An explicit alignment may be specified for a global. If not present, or if
778 the alignment is set to zero, the alignment of the global is set by the target
779 to whatever it feels convenient. If an explicit alignment is specified, the
780 global is forced to have at least that much alignment. All alignments must be
783 <p>For example, the following defines a global in a numbered address space with
784 an initializer, section, and alignment:</p>
786 <div class="doc_code">
788 @G = addrspace(5) constant float 1.0, section "foo", align 4
795 <!-- ======================================================================= -->
796 <div class="doc_subsection">
797 <a name="functionstructure">Functions</a>
800 <div class="doc_text">
802 <p>LLVM function definitions consist of the "<tt>define</tt>" keyord,
803 an optional <a href="#linkage">linkage type</a>, an optional
804 <a href="#visibility">visibility style</a>, an optional
805 <a href="#callingconv">calling convention</a>, a return type, an optional
806 <a href="#paramattrs">parameter attribute</a> for the return type, a function
807 name, a (possibly empty) argument list (each with optional
808 <a href="#paramattrs">parameter attributes</a>), optional
809 <a href="#fnattrs">function attributes</a>, an optional section,
810 an optional alignment, an optional <a href="#gc">garbage collector name</a>,
811 an opening curly brace, a list of basic blocks, and a closing curly brace.
813 LLVM function declarations consist of the "<tt>declare</tt>" keyword, an
814 optional <a href="#linkage">linkage type</a>, an optional
815 <a href="#visibility">visibility style</a>, an optional
816 <a href="#callingconv">calling convention</a>, a return type, an optional
817 <a href="#paramattrs">parameter attribute</a> for the return type, a function
818 name, a possibly empty list of arguments, an optional alignment, and an optional
819 <a href="#gc">garbage collector name</a>.</p>
821 <p>A function definition contains a list of basic blocks, forming the CFG
822 (Control Flow Graph) for
823 the function. Each basic block may optionally start with a label (giving the
824 basic block a symbol table entry), contains a list of instructions, and ends
825 with a <a href="#terminators">terminator</a> instruction (such as a branch or
826 function return).</p>
828 <p>The first basic block in a function is special in two ways: it is immediately
829 executed on entrance to the function, and it is not allowed to have predecessor
830 basic blocks (i.e. there can not be any branches to the entry block of a
831 function). Because the block can have no predecessors, it also cannot have any
832 <a href="#i_phi">PHI nodes</a>.</p>
834 <p>LLVM allows an explicit section to be specified for functions. If the target
835 supports it, it will emit functions to the section specified.</p>
837 <p>An explicit alignment may be specified for a function. If not present, or if
838 the alignment is set to zero, the alignment of the function is set by the target
839 to whatever it feels convenient. If an explicit alignment is specified, the
840 function is forced to have at least that much alignment. All alignments must be
845 <div class="doc_code">
847 define [<a href="#linkage">linkage</a>] [<a href="#visibility">visibility</a>]
848 [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>]
849 <ResultType> @<FunctionName> ([argument list])
850 [<a href="#fnattrs">fn Attrs</a>] [section "name"] [align N]
851 [<a href="#gc">gc</a>] { ... }
858 <!-- ======================================================================= -->
859 <div class="doc_subsection">
860 <a name="aliasstructure">Aliases</a>
862 <div class="doc_text">
863 <p>Aliases act as "second name" for the aliasee value (which can be either
864 function, global variable, another alias or bitcast of global value). Aliases
865 may have an optional <a href="#linkage">linkage type</a>, and an
866 optional <a href="#visibility">visibility style</a>.</p>
870 <div class="doc_code">
872 @<Name> = alias [Linkage] [Visibility] <AliaseeTy> @<Aliasee>
880 <!-- ======================================================================= -->
881 <div class="doc_subsection"><a name="paramattrs">Parameter Attributes</a></div>
882 <div class="doc_text">
883 <p>The return type and each parameter of a function type may have a set of
884 <i>parameter attributes</i> associated with them. Parameter attributes are
885 used to communicate additional information about the result or parameters of
886 a function. Parameter attributes are considered to be part of the function,
887 not of the function type, so functions with different parameter attributes
888 can have the same function type.</p>
890 <p>Parameter attributes are simple keywords that follow the type specified. If
891 multiple parameter attributes are needed, they are space separated. For
894 <div class="doc_code">
896 declare i32 @printf(i8* noalias , ...)
897 declare i32 @atoi(i8 zeroext)
898 declare signext i8 @returns_signed_char()
902 <p>Note that any attributes for the function result (<tt>nounwind</tt>,
903 <tt>readonly</tt>) come immediately after the argument list.</p>
905 <p>Currently, only the following parameter attributes are defined:</p>
907 <dt><tt>zeroext</tt></dt>
908 <dd>This indicates to the code generator that the parameter or return value
909 should be zero-extended to a 32-bit value by the caller (for a parameter)
910 or the callee (for a return value).</dd>
912 <dt><tt>signext</tt></dt>
913 <dd>This indicates to the code generator that the parameter or return value
914 should be sign-extended to a 32-bit value by the caller (for a parameter)
915 or the callee (for a return value).</dd>
917 <dt><tt>inreg</tt></dt>
918 <dd>This indicates that this parameter or return value should be treated
919 in a special target-dependent fashion during while emitting code for a
920 function call or return (usually, by putting it in a register as opposed
921 to memory, though some targets use it to distinguish between two different
922 kinds of registers). Use of this attribute is target-specific.</dd>
924 <dt><tt><a name="byval">byval</a></tt></dt>
925 <dd>This indicates that the pointer parameter should really be passed by
926 value to the function. The attribute implies that a hidden copy of the
927 pointee is made between the caller and the callee, so the callee is unable
928 to modify the value in the callee. This attribute is only valid on LLVM
929 pointer arguments. It is generally used to pass structs and arrays by
930 value, but is also valid on pointers to scalars. The copy is considered to
931 belong to the caller not the callee (for example,
932 <tt><a href="#readonly">readonly</a></tt> functions should not write to
933 <tt>byval</tt> parameters). This is not a valid attribute for return
934 values. The byval attribute also supports specifying an alignment with the
935 align attribute. This has a target-specific effect on the code generator
936 that usually indicates a desired alignment for the synthesized stack
939 <dt><tt>sret</tt></dt>
940 <dd>This indicates that the pointer parameter specifies the address of a
941 structure that is the return value of the function in the source program.
942 This pointer must be guaranteed by the caller to be valid: loads and stores
943 to the structure may be assumed by the callee to not to trap. This may only
944 be applied to the first parameter. This is not a valid attribute for
947 <dt><tt>noalias</tt></dt>
948 <dd>This indicates that the pointer does not alias any global or any other
949 parameter. The caller is responsible for ensuring that this is the
950 case. On a function return value, <tt>noalias</tt> additionally indicates
951 that the pointer does not alias any other pointers visible to the
952 caller. For further details, please see the discussion of the NoAlias
954 <a href="http://llvm.org/docs/AliasAnalysis.html#MustMayNo">alias
957 <dt><tt>nocapture</tt></dt>
958 <dd>This indicates that the callee does not make any copies of the pointer
959 that outlive the callee itself. This is not a valid attribute for return
962 <dt><tt>nest</tt></dt>
963 <dd>This indicates that the pointer parameter can be excised using the
964 <a href="#int_trampoline">trampoline intrinsics</a>. This is not a valid
965 attribute for return values.</dd>
970 <!-- ======================================================================= -->
971 <div class="doc_subsection">
972 <a name="gc">Garbage Collector Names</a>
975 <div class="doc_text">
976 <p>Each function may specify a garbage collector name, which is simply a
979 <div class="doc_code"><pre
980 >define void @f() gc "name" { ...</pre></div>
982 <p>The compiler declares the supported values of <i>name</i>. Specifying a
983 collector which will cause the compiler to alter its output in order to support
984 the named garbage collection algorithm.</p>
987 <!-- ======================================================================= -->
988 <div class="doc_subsection">
989 <a name="fnattrs">Function Attributes</a>
992 <div class="doc_text">
994 <p>Function attributes are set to communicate additional information about
995 a function. Function attributes are considered to be part of the function,
996 not of the function type, so functions with different parameter attributes
997 can have the same function type.</p>
999 <p>Function attributes are simple keywords that follow the type specified. If
1000 multiple attributes are needed, they are space separated. For
1003 <div class="doc_code">
1005 define void @f() noinline { ... }
1006 define void @f() alwaysinline { ... }
1007 define void @f() alwaysinline optsize { ... }
1008 define void @f() optsize
1013 <dt><tt>alwaysinline</tt></dt>
1014 <dd>This attribute indicates that the inliner should attempt to inline this
1015 function into callers whenever possible, ignoring any active inlining size
1016 threshold for this caller.</dd>
1018 <dt><tt>noinline</tt></dt>
1019 <dd>This attribute indicates that the inliner should never inline this function
1020 in any situation. This attribute may not be used together with the
1021 <tt>alwaysinline</tt> attribute.</dd>
1023 <dt><tt>optsize</tt></dt>
1024 <dd>This attribute suggests that optimization passes and code generator passes
1025 make choices that keep the code size of this function low, and otherwise do
1026 optimizations specifically to reduce code size.</dd>
1028 <dt><tt>noreturn</tt></dt>
1029 <dd>This function attribute indicates that the function never returns normally.
1030 This produces undefined behavior at runtime if the function ever does
1031 dynamically return.</dd>
1033 <dt><tt>nounwind</tt></dt>
1034 <dd>This function attribute indicates that the function never returns with an
1035 unwind or exceptional control flow. If the function does unwind, its runtime
1036 behavior is undefined.</dd>
1038 <dt><tt>readnone</tt></dt>
1039 <dd>This attribute indicates that the function computes its result (or the
1040 exception it throws) based strictly on its arguments, without dereferencing any
1041 pointer arguments or otherwise accessing any mutable state (e.g. memory, control
1042 registers, etc) visible to caller functions. It does not write through any
1043 pointer arguments (including <tt><a href="#byval">byval</a></tt> arguments) and
1044 never changes any state visible to callers.</dd>
1046 <dt><tt><a name="readonly">readonly</a></tt></dt>
1047 <dd>This attribute indicates that the function does not write through any
1048 pointer arguments (including <tt><a href="#byval">byval</a></tt> arguments)
1049 or otherwise modify any state (e.g. memory, control registers, etc) visible to
1050 caller functions. It may dereference pointer arguments and read state that may
1051 be set in the caller. A readonly function always returns the same value (or
1052 throws the same exception) when called with the same set of arguments and global
1055 <dt><tt><a name="ssp">ssp</a></tt></dt>
1056 <dd>This attribute indicates that the function should emit a stack smashing
1057 protector. It is in the form of a "canary"—a random value placed on the
1058 stack before the local variables that's checked upon return from the function to
1059 see if it has been overwritten. A heuristic is used to determine if a function
1060 needs stack protectors or not.
1062 <p>If a function that has an <tt>ssp</tt> attribute is inlined into a function
1063 that doesn't have an <tt>ssp</tt> attribute, then the resulting function will
1064 have an <tt>ssp</tt> attribute.</p></dd>
1066 <dt><tt>sspreq</tt></dt>
1067 <dd>This attribute indicates that the function should <em>always</em> emit a
1068 stack smashing protector. This overrides the <tt><a href="#ssp">ssp</a></tt>
1071 <p>If a function that has an <tt>sspreq</tt> attribute is inlined into a
1072 function that doesn't have an <tt>sspreq</tt> attribute or which has
1073 an <tt>ssp</tt> attribute, then the resulting function will have
1074 an <tt>sspreq</tt> attribute.</p></dd>
1079 <!-- ======================================================================= -->
1080 <div class="doc_subsection">
1081 <a name="moduleasm">Module-Level Inline Assembly</a>
1084 <div class="doc_text">
1086 Modules may contain "module-level inline asm" blocks, which corresponds to the
1087 GCC "file scope inline asm" blocks. These blocks are internally concatenated by
1088 LLVM and treated as a single unit, but may be separated in the .ll file if
1089 desired. The syntax is very simple:
1092 <div class="doc_code">
1094 module asm "inline asm code goes here"
1095 module asm "more can go here"
1099 <p>The strings can contain any character by escaping non-printable characters.
1100 The escape sequence used is simply "\xx" where "xx" is the two digit hex code
1105 The inline asm code is simply printed to the machine code .s file when
1106 assembly code is generated.
1110 <!-- ======================================================================= -->
1111 <div class="doc_subsection">
1112 <a name="datalayout">Data Layout</a>
1115 <div class="doc_text">
1116 <p>A module may specify a target specific data layout string that specifies how
1117 data is to be laid out in memory. The syntax for the data layout is simply:</p>
1118 <pre> target datalayout = "<i>layout specification</i>"</pre>
1119 <p>The <i>layout specification</i> consists of a list of specifications
1120 separated by the minus sign character ('-'). Each specification starts with a
1121 letter and may include other information after the letter to define some
1122 aspect of the data layout. The specifications accepted are as follows: </p>
1125 <dd>Specifies that the target lays out data in big-endian form. That is, the
1126 bits with the most significance have the lowest address location.</dd>
1128 <dd>Specifies that the target lays out data in little-endian form. That is,
1129 the bits with the least significance have the lowest address location.</dd>
1130 <dt><tt>p:<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1131 <dd>This specifies the <i>size</i> of a pointer and its <i>abi</i> and
1132 <i>preferred</i> alignments. All sizes are in bits. Specifying the <i>pref</i>
1133 alignment is optional. If omitted, the preceding <tt>:</tt> should be omitted
1135 <dt><tt>i<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1136 <dd>This specifies the alignment for an integer type of a given bit
1137 <i>size</i>. The value of <i>size</i> must be in the range [1,2^23).</dd>
1138 <dt><tt>v<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1139 <dd>This specifies the alignment for a vector type of a given bit
1141 <dt><tt>f<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1142 <dd>This specifies the alignment for a floating point type of a given bit
1143 <i>size</i>. The value of <i>size</i> must be either 32 (float) or 64
1145 <dt><tt>a<i>size</i>:<i>abi</i>:<i>pref</i></tt></dt>
1146 <dd>This specifies the alignment for an aggregate type of a given bit
1149 <p>When constructing the data layout for a given target, LLVM starts with a
1150 default set of specifications which are then (possibly) overriden by the
1151 specifications in the <tt>datalayout</tt> keyword. The default specifications
1152 are given in this list:</p>
1154 <li><tt>E</tt> - big endian</li>
1155 <li><tt>p:32:64:64</tt> - 32-bit pointers with 64-bit alignment</li>
1156 <li><tt>i1:8:8</tt> - i1 is 8-bit (byte) aligned</li>
1157 <li><tt>i8:8:8</tt> - i8 is 8-bit (byte) aligned</li>
1158 <li><tt>i16:16:16</tt> - i16 is 16-bit aligned</li>
1159 <li><tt>i32:32:32</tt> - i32 is 32-bit aligned</li>
1160 <li><tt>i64:32:64</tt> - i64 has ABI alignment of 32-bits but preferred
1161 alignment of 64-bits</li>
1162 <li><tt>f32:32:32</tt> - float is 32-bit aligned</li>
1163 <li><tt>f64:64:64</tt> - double is 64-bit aligned</li>
1164 <li><tt>v64:64:64</tt> - 64-bit vector is 64-bit aligned</li>
1165 <li><tt>v128:128:128</tt> - 128-bit vector is 128-bit aligned</li>
1166 <li><tt>a0:0:1</tt> - aggregates are 8-bit aligned</li>
1168 <p>When LLVM is determining the alignment for a given type, it uses the
1169 following rules:</p>
1171 <li>If the type sought is an exact match for one of the specifications, that
1172 specification is used.</li>
1173 <li>If no match is found, and the type sought is an integer type, then the
1174 smallest integer type that is larger than the bitwidth of the sought type is
1175 used. If none of the specifications are larger than the bitwidth then the the
1176 largest integer type is used. For example, given the default specifications
1177 above, the i7 type will use the alignment of i8 (next largest) while both
1178 i65 and i256 will use the alignment of i64 (largest specified).</li>
1179 <li>If no match is found, and the type sought is a vector type, then the
1180 largest vector type that is smaller than the sought vector type will be used
1181 as a fall back. This happens because <128 x double> can be implemented
1182 in terms of 64 <2 x double>, for example.</li>
1186 <!-- *********************************************************************** -->
1187 <div class="doc_section"> <a name="typesystem">Type System</a> </div>
1188 <!-- *********************************************************************** -->
1190 <div class="doc_text">
1192 <p>The LLVM type system is one of the most important features of the
1193 intermediate representation. Being typed enables a number of
1194 optimizations to be performed on the intermediate representation directly,
1195 without having to do
1196 extra analyses on the side before the transformation. A strong type
1197 system makes it easier to read the generated code and enables novel
1198 analyses and transformations that are not feasible to perform on normal
1199 three address code representations.</p>
1203 <!-- ======================================================================= -->
1204 <div class="doc_subsection"> <a name="t_classifications">Type
1205 Classifications</a> </div>
1206 <div class="doc_text">
1207 <p>The types fall into a few useful
1208 classifications:</p>
1210 <table border="1" cellspacing="0" cellpadding="4">
1212 <tr><th>Classification</th><th>Types</th></tr>
1214 <td><a href="#t_integer">integer</a></td>
1215 <td><tt>i1, i2, i3, ... i8, ... i16, ... i32, ... i64, ... </tt></td>
1218 <td><a href="#t_floating">floating point</a></td>
1219 <td><tt>float, double, x86_fp80, fp128, ppc_fp128</tt></td>
1222 <td><a name="t_firstclass">first class</a></td>
1223 <td><a href="#t_integer">integer</a>,
1224 <a href="#t_floating">floating point</a>,
1225 <a href="#t_pointer">pointer</a>,
1226 <a href="#t_vector">vector</a>,
1227 <a href="#t_struct">structure</a>,
1228 <a href="#t_array">array</a>,
1229 <a href="#t_label">label</a>.
1233 <td><a href="#t_primitive">primitive</a></td>
1234 <td><a href="#t_label">label</a>,
1235 <a href="#t_void">void</a>,
1236 <a href="#t_floating">floating point</a>.</td>
1239 <td><a href="#t_derived">derived</a></td>
1240 <td><a href="#t_integer">integer</a>,
1241 <a href="#t_array">array</a>,
1242 <a href="#t_function">function</a>,
1243 <a href="#t_pointer">pointer</a>,
1244 <a href="#t_struct">structure</a>,
1245 <a href="#t_pstruct">packed structure</a>,
1246 <a href="#t_vector">vector</a>,
1247 <a href="#t_opaque">opaque</a>.
1253 <p>The <a href="#t_firstclass">first class</a> types are perhaps the
1254 most important. Values of these types are the only ones which can be
1255 produced by instructions, passed as arguments, or used as operands to
1259 <!-- ======================================================================= -->
1260 <div class="doc_subsection"> <a name="t_primitive">Primitive Types</a> </div>
1262 <div class="doc_text">
1263 <p>The primitive types are the fundamental building blocks of the LLVM
1268 <!-- _______________________________________________________________________ -->
1269 <div class="doc_subsubsection"> <a name="t_floating">Floating Point Types</a> </div>
1271 <div class="doc_text">
1274 <tr><th>Type</th><th>Description</th></tr>
1275 <tr><td><tt>float</tt></td><td>32-bit floating point value</td></tr>
1276 <tr><td><tt>double</tt></td><td>64-bit floating point value</td></tr>
1277 <tr><td><tt>fp128</tt></td><td>128-bit floating point value (112-bit mantissa)</td></tr>
1278 <tr><td><tt>x86_fp80</tt></td><td>80-bit floating point value (X87)</td></tr>
1279 <tr><td><tt>ppc_fp128</tt></td><td>128-bit floating point value (two 64-bits)</td></tr>
1284 <!-- _______________________________________________________________________ -->
1285 <div class="doc_subsubsection"> <a name="t_void">Void Type</a> </div>
1287 <div class="doc_text">
1289 <p>The void type does not represent any value and has no size.</p>
1298 <!-- _______________________________________________________________________ -->
1299 <div class="doc_subsubsection"> <a name="t_label">Label Type</a> </div>
1301 <div class="doc_text">
1303 <p>The label type represents code labels.</p>
1313 <!-- ======================================================================= -->
1314 <div class="doc_subsection"> <a name="t_derived">Derived Types</a> </div>
1316 <div class="doc_text">
1318 <p>The real power in LLVM comes from the derived types in the system.
1319 This is what allows a programmer to represent arrays, functions,
1320 pointers, and other useful types. Note that these derived types may be
1321 recursive: For example, it is possible to have a two dimensional array.</p>
1325 <!-- _______________________________________________________________________ -->
1326 <div class="doc_subsubsection"> <a name="t_integer">Integer Type</a> </div>
1328 <div class="doc_text">
1331 <p>The integer type is a very simple derived type that simply specifies an
1332 arbitrary bit width for the integer type desired. Any bit width from 1 bit to
1333 2^23-1 (about 8 million) can be specified.</p>
1341 <p>The number of bits the integer will occupy is specified by the <tt>N</tt>
1345 <table class="layout">
1348 <td><tt>i1</tt></td>
1349 <td>a single-bit integer.</td>
1351 <td><tt>i32</tt></td>
1352 <td>a 32-bit integer.</td>
1354 <td><tt>i1942652</tt></td>
1355 <td>a really big integer of over 1 million bits.</td>
1360 <p>Note that the code generator does not yet support large integer types
1361 to be used as function return types. The specific limit on how large a
1362 return type the code generator can currently handle is target-dependent;
1363 currently it's often 64 bits for 32-bit targets and 128 bits for 64-bit
1368 <!-- _______________________________________________________________________ -->
1369 <div class="doc_subsubsection"> <a name="t_array">Array Type</a> </div>
1371 <div class="doc_text">
1375 <p>The array type is a very simple derived type that arranges elements
1376 sequentially in memory. The array type requires a size (number of
1377 elements) and an underlying data type.</p>
1382 [<# elements> x <elementtype>]
1385 <p>The number of elements is a constant integer value; elementtype may
1386 be any type with a size.</p>
1389 <table class="layout">
1391 <td class="left"><tt>[40 x i32]</tt></td>
1392 <td class="left">Array of 40 32-bit integer values.</td>
1395 <td class="left"><tt>[41 x i32]</tt></td>
1396 <td class="left">Array of 41 32-bit integer values.</td>
1399 <td class="left"><tt>[4 x i8]</tt></td>
1400 <td class="left">Array of 4 8-bit integer values.</td>
1403 <p>Here are some examples of multidimensional arrays:</p>
1404 <table class="layout">
1406 <td class="left"><tt>[3 x [4 x i32]]</tt></td>
1407 <td class="left">3x4 array of 32-bit integer values.</td>
1410 <td class="left"><tt>[12 x [10 x float]]</tt></td>
1411 <td class="left">12x10 array of single precision floating point values.</td>
1414 <td class="left"><tt>[2 x [3 x [4 x i16]]]</tt></td>
1415 <td class="left">2x3x4 array of 16-bit integer values.</td>
1419 <p>Note that 'variable sized arrays' can be implemented in LLVM with a zero
1420 length array. Normally, accesses past the end of an array are undefined in
1421 LLVM (e.g. it is illegal to access the 5th element of a 3 element array).
1422 As a special case, however, zero length arrays are recognized to be variable
1423 length. This allows implementation of 'pascal style arrays' with the LLVM
1424 type "{ i32, [0 x float]}", for example.</p>
1426 <p>Note that the code generator does not yet support large aggregate types
1427 to be used as function return types. The specific limit on how large an
1428 aggregate return type the code generator can currently handle is
1429 target-dependent, and also dependent on the aggregate element types.</p>
1433 <!-- _______________________________________________________________________ -->
1434 <div class="doc_subsubsection"> <a name="t_function">Function Type</a> </div>
1435 <div class="doc_text">
1439 <p>The function type can be thought of as a function signature. It
1440 consists of a return type and a list of formal parameter types. The
1441 return type of a function type is a scalar type, a void type, or a struct type.
1442 If the return type is a struct type then all struct elements must be of first
1443 class types, and the struct must have at least one element.</p>
1448 <returntype list> (<parameter list>)
1451 <p>...where '<tt><parameter list></tt>' is a comma-separated list of type
1452 specifiers. Optionally, the parameter list may include a type <tt>...</tt>,
1453 which indicates that the function takes a variable number of arguments.
1454 Variable argument functions can access their arguments with the <a
1455 href="#int_varargs">variable argument handling intrinsic</a> functions.
1456 '<tt><returntype list></tt>' is a comma-separated list of
1457 <a href="#t_firstclass">first class</a> type specifiers.</p>
1460 <table class="layout">
1462 <td class="left"><tt>i32 (i32)</tt></td>
1463 <td class="left">function taking an <tt>i32</tt>, returning an <tt>i32</tt>
1465 </tr><tr class="layout">
1466 <td class="left"><tt>float (i16 signext, i32 *) *
1468 <td class="left"><a href="#t_pointer">Pointer</a> to a function that takes
1469 an <tt>i16</tt> that should be sign extended and a
1470 <a href="#t_pointer">pointer</a> to <tt>i32</tt>, returning
1473 </tr><tr class="layout">
1474 <td class="left"><tt>i32 (i8*, ...)</tt></td>
1475 <td class="left">A vararg function that takes at least one
1476 <a href="#t_pointer">pointer</a> to <tt>i8 </tt> (char in C),
1477 which returns an integer. This is the signature for <tt>printf</tt> in
1480 </tr><tr class="layout">
1481 <td class="left"><tt>{i32, i32} (i32)</tt></td>
1482 <td class="left">A function taking an <tt>i32</tt>, returning two
1483 <tt>i32</tt> values as an aggregate of type <tt>{ i32, i32 }</tt>
1489 <!-- _______________________________________________________________________ -->
1490 <div class="doc_subsubsection"> <a name="t_struct">Structure Type</a> </div>
1491 <div class="doc_text">
1493 <p>The structure type is used to represent a collection of data members
1494 together in memory. The packing of the field types is defined to match
1495 the ABI of the underlying processor. The elements of a structure may
1496 be any type that has a size.</p>
1497 <p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1498 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1499 field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1502 <pre> { <type list> }<br></pre>
1504 <table class="layout">
1506 <td class="left"><tt>{ i32, i32, i32 }</tt></td>
1507 <td class="left">A triple of three <tt>i32</tt> values</td>
1508 </tr><tr class="layout">
1509 <td class="left"><tt>{ float, i32 (i32) * }</tt></td>
1510 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1511 second element is a <a href="#t_pointer">pointer</a> to a
1512 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1513 an <tt>i32</tt>.</td>
1517 <p>Note that the code generator does not yet support large aggregate types
1518 to be used as function return types. The specific limit on how large an
1519 aggregate return type the code generator can currently handle is
1520 target-dependent, and also dependent on the aggregate element types.</p>
1524 <!-- _______________________________________________________________________ -->
1525 <div class="doc_subsubsection"> <a name="t_pstruct">Packed Structure Type</a>
1527 <div class="doc_text">
1529 <p>The packed structure type is used to represent a collection of data members
1530 together in memory. There is no padding between fields. Further, the alignment
1531 of a packed structure is 1 byte. The elements of a packed structure may
1532 be any type that has a size.</p>
1533 <p>Structures are accessed using '<tt><a href="#i_load">load</a></tt>
1534 and '<tt><a href="#i_store">store</a></tt>' by getting a pointer to a
1535 field with the '<tt><a href="#i_getelementptr">getelementptr</a></tt>'
1538 <pre> < { <type list> } > <br></pre>
1540 <table class="layout">
1542 <td class="left"><tt>< { i32, i32, i32 } ></tt></td>
1543 <td class="left">A triple of three <tt>i32</tt> values</td>
1544 </tr><tr class="layout">
1546 <tt>< { float, i32 (i32)* } ></tt></td>
1547 <td class="left">A pair, where the first element is a <tt>float</tt> and the
1548 second element is a <a href="#t_pointer">pointer</a> to a
1549 <a href="#t_function">function</a> that takes an <tt>i32</tt>, returning
1550 an <tt>i32</tt>.</td>
1555 <!-- _______________________________________________________________________ -->
1556 <div class="doc_subsubsection"> <a name="t_pointer">Pointer Type</a> </div>
1557 <div class="doc_text">
1559 <p>As in many languages, the pointer type represents a pointer or
1560 reference to another object, which must live in memory. Pointer types may have
1561 an optional address space attribute defining the target-specific numbered
1562 address space where the pointed-to object resides. The default address space is
1565 <p>Note that LLVM does not permit pointers to void (<tt>void*</tt>) nor does
1566 it permit pointers to labels (<tt>label*</tt>). Use <tt>i8*</tt> instead.</p>
1569 <pre> <type> *<br></pre>
1571 <table class="layout">
1573 <td class="left"><tt>[4 x i32]*</tt></td>
1574 <td class="left">A <a href="#t_pointer">pointer</a> to <a
1575 href="#t_array">array</a> of four <tt>i32</tt> values.</td>
1578 <td class="left"><tt>i32 (i32 *) *</tt></td>
1579 <td class="left"> A <a href="#t_pointer">pointer</a> to a <a
1580 href="#t_function">function</a> that takes an <tt>i32*</tt>, returning an
1584 <td class="left"><tt>i32 addrspace(5)*</tt></td>
1585 <td class="left">A <a href="#t_pointer">pointer</a> to an <tt>i32</tt> value
1586 that resides in address space #5.</td>
1591 <!-- _______________________________________________________________________ -->
1592 <div class="doc_subsubsection"> <a name="t_vector">Vector Type</a> </div>
1593 <div class="doc_text">
1597 <p>A vector type is a simple derived type that represents a vector
1598 of elements. Vector types are used when multiple primitive data
1599 are operated in parallel using a single instruction (SIMD).
1600 A vector type requires a size (number of
1601 elements) and an underlying primitive data type. Vectors must have a power
1602 of two length (1, 2, 4, 8, 16 ...). Vector types are
1603 considered <a href="#t_firstclass">first class</a>.</p>
1608 < <# elements> x <elementtype> >
1611 <p>The number of elements is a constant integer value; elementtype may
1612 be any integer or floating point type.</p>
1616 <table class="layout">
1618 <td class="left"><tt><4 x i32></tt></td>
1619 <td class="left">Vector of 4 32-bit integer values.</td>
1622 <td class="left"><tt><8 x float></tt></td>
1623 <td class="left">Vector of 8 32-bit floating-point values.</td>
1626 <td class="left"><tt><2 x i64></tt></td>
1627 <td class="left">Vector of 2 64-bit integer values.</td>
1631 <p>Note that the code generator does not yet support large vector types
1632 to be used as function return types. The specific limit on how large a
1633 vector return type codegen can currently handle is target-dependent;
1634 currently it's often a few times longer than a hardware vector register.</p>
1638 <!-- _______________________________________________________________________ -->
1639 <div class="doc_subsubsection"> <a name="t_opaque">Opaque Type</a> </div>
1640 <div class="doc_text">
1644 <p>Opaque types are used to represent unknown types in the system. This
1645 corresponds (for example) to the C notion of a forward declared structure type.
1646 In LLVM, opaque types can eventually be resolved to any type (not just a
1647 structure type).</p>
1657 <table class="layout">
1659 <td class="left"><tt>opaque</tt></td>
1660 <td class="left">An opaque type.</td>
1665 <!-- ======================================================================= -->
1666 <div class="doc_subsection">
1667 <a name="t_uprefs">Type Up-references</a>
1670 <div class="doc_text">
1673 An "up reference" allows you to refer to a lexically enclosing type without
1674 requiring it to have a name. For instance, a structure declaration may contain a
1675 pointer to any of the types it is lexically a member of. Example of up
1676 references (with their equivalent as named type declarations) include:</p>
1679 { \2 * } %x = type { %t* }
1680 { \2 }* %y = type { %y }*
1685 An up reference is needed by the asmprinter for printing out cyclic types when
1686 there is no declared name for a type in the cycle. Because the asmprinter does
1687 not want to print out an infinite type string, it needs a syntax to handle
1688 recursive types that have no names (all names are optional in llvm IR).
1697 The level is the count of the lexical type that is being referred to.
1702 <table class="layout">
1704 <td class="left"><tt>\1*</tt></td>
1705 <td class="left">Self-referential pointer.</td>
1708 <td class="left"><tt>{ { \3*, i8 }, i32 }</tt></td>
1709 <td class="left">Recursive structure where the upref refers to the out-most
1716 <!-- *********************************************************************** -->
1717 <div class="doc_section"> <a name="constants">Constants</a> </div>
1718 <!-- *********************************************************************** -->
1720 <div class="doc_text">
1722 <p>LLVM has several different basic types of constants. This section describes
1723 them all and their syntax.</p>
1727 <!-- ======================================================================= -->
1728 <div class="doc_subsection"><a name="simpleconstants">Simple Constants</a></div>
1730 <div class="doc_text">
1733 <dt><b>Boolean constants</b></dt>
1735 <dd>The two strings '<tt>true</tt>' and '<tt>false</tt>' are both valid
1736 constants of the <tt><a href="#t_primitive">i1</a></tt> type.
1739 <dt><b>Integer constants</b></dt>
1741 <dd>Standard integers (such as '4') are constants of the <a
1742 href="#t_integer">integer</a> type. Negative numbers may be used with
1746 <dt><b>Floating point constants</b></dt>
1748 <dd>Floating point constants use standard decimal notation (e.g. 123.421),
1749 exponential notation (e.g. 1.23421e+2), or a more precise hexadecimal
1750 notation (see below). The assembler requires the exact decimal value of
1751 a floating-point constant. For example, the assembler accepts 1.25 but
1752 rejects 1.3 because 1.3 is a repeating decimal in binary. Floating point
1753 constants must have a <a href="#t_floating">floating point</a> type. </dd>
1755 <dt><b>Null pointer constants</b></dt>
1757 <dd>The identifier '<tt>null</tt>' is recognized as a null pointer constant
1758 and must be of <a href="#t_pointer">pointer type</a>.</dd>
1762 <p>The one non-intuitive notation for constants is the optional hexadecimal form
1763 of floating point constants. For example, the form '<tt>double
1764 0x432ff973cafa8000</tt>' is equivalent to (but harder to read than) '<tt>double
1765 4.5e+15</tt>'. The only time hexadecimal floating point constants are required
1766 (and the only time that they are generated by the disassembler) is when a
1767 floating point constant must be emitted but it cannot be represented as a
1768 decimal floating point number. For example, NaN's, infinities, and other
1769 special values are represented in their IEEE hexadecimal format so that
1770 assembly and disassembly do not cause any bits to change in the constants.</p>
1774 <!-- ======================================================================= -->
1775 <div class="doc_subsection"><a name="aggregateconstants">Aggregate Constants</a>
1778 <div class="doc_text">
1779 <p>Aggregate constants arise from aggregation of simple constants
1780 and smaller aggregate constants.</p>
1783 <dt><b>Structure constants</b></dt>
1785 <dd>Structure constants are represented with notation similar to structure
1786 type definitions (a comma separated list of elements, surrounded by braces
1787 (<tt>{}</tt>)). For example: "<tt>{ i32 4, float 17.0, i32* @G }</tt>",
1788 where "<tt>@G</tt>" is declared as "<tt>@G = external global i32</tt>". Structure constants
1789 must have <a href="#t_struct">structure type</a>, and the number and
1790 types of elements must match those specified by the type.
1793 <dt><b>Array constants</b></dt>
1795 <dd>Array constants are represented with notation similar to array type
1796 definitions (a comma separated list of elements, surrounded by square brackets
1797 (<tt>[]</tt>)). For example: "<tt>[ i32 42, i32 11, i32 74 ]</tt>". Array
1798 constants must have <a href="#t_array">array type</a>, and the number and
1799 types of elements must match those specified by the type.
1802 <dt><b>Vector constants</b></dt>
1804 <dd>Vector constants are represented with notation similar to vector type
1805 definitions (a comma separated list of elements, surrounded by
1806 less-than/greater-than's (<tt><></tt>)). For example: "<tt>< i32 42,
1807 i32 11, i32 74, i32 100 ></tt>". Vector constants must have <a
1808 href="#t_vector">vector type</a>, and the number and types of elements must
1809 match those specified by the type.
1812 <dt><b>Zero initialization</b></dt>
1814 <dd>The string '<tt>zeroinitializer</tt>' can be used to zero initialize a
1815 value to zero of <em>any</em> type, including scalar and aggregate types.
1816 This is often used to avoid having to print large zero initializers (e.g. for
1817 large arrays) and is always exactly equivalent to using explicit zero
1824 <!-- ======================================================================= -->
1825 <div class="doc_subsection">
1826 <a name="globalconstants">Global Variable and Function Addresses</a>
1829 <div class="doc_text">
1831 <p>The addresses of <a href="#globalvars">global variables</a> and <a
1832 href="#functionstructure">functions</a> are always implicitly valid (link-time)
1833 constants. These constants are explicitly referenced when the <a
1834 href="#identifiers">identifier for the global</a> is used and always have <a
1835 href="#t_pointer">pointer</a> type. For example, the following is a legal LLVM
1838 <div class="doc_code">
1842 @Z = global [2 x i32*] [ i32* @X, i32* @Y ]
1848 <!-- ======================================================================= -->
1849 <div class="doc_subsection"><a name="undefvalues">Undefined Values</a></div>
1850 <div class="doc_text">
1851 <p>The string '<tt>undef</tt>' is recognized as a type-less constant that has
1852 no specific value. Undefined values may be of any type and be used anywhere
1853 a constant is permitted.</p>
1855 <p>Undefined values indicate to the compiler that the program is well defined
1856 no matter what value is used, giving the compiler more freedom to optimize.
1860 <!-- ======================================================================= -->
1861 <div class="doc_subsection"><a name="constantexprs">Constant Expressions</a>
1864 <div class="doc_text">
1866 <p>Constant expressions are used to allow expressions involving other constants
1867 to be used as constants. Constant expressions may be of any <a
1868 href="#t_firstclass">first class</a> type and may involve any LLVM operation
1869 that does not have side effects (e.g. load and call are not supported). The
1870 following is the syntax for constant expressions:</p>
1873 <dt><b><tt>trunc ( CST to TYPE )</tt></b></dt>
1874 <dd>Truncate a constant to another type. The bit size of CST must be larger
1875 than the bit size of TYPE. Both types must be integers.</dd>
1877 <dt><b><tt>zext ( CST to TYPE )</tt></b></dt>
1878 <dd>Zero extend a constant to another type. The bit size of CST must be
1879 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1881 <dt><b><tt>sext ( CST to TYPE )</tt></b></dt>
1882 <dd>Sign extend a constant to another type. The bit size of CST must be
1883 smaller or equal to the bit size of TYPE. Both types must be integers.</dd>
1885 <dt><b><tt>fptrunc ( CST to TYPE )</tt></b></dt>
1886 <dd>Truncate a floating point constant to another floating point type. The
1887 size of CST must be larger than the size of TYPE. Both types must be
1888 floating point.</dd>
1890 <dt><b><tt>fpext ( CST to TYPE )</tt></b></dt>
1891 <dd>Floating point extend a constant to another type. The size of CST must be
1892 smaller or equal to the size of TYPE. Both types must be floating point.</dd>
1894 <dt><b><tt>fptoui ( CST to TYPE )</tt></b></dt>
1895 <dd>Convert a floating point constant to the corresponding unsigned integer
1896 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1897 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1898 of the same number of elements. If the value won't fit in the integer type,
1899 the results are undefined.</dd>
1901 <dt><b><tt>fptosi ( CST to TYPE )</tt></b></dt>
1902 <dd>Convert a floating point constant to the corresponding signed integer
1903 constant. TYPE must be a scalar or vector integer type. CST must be of scalar
1904 or vector floating point type. Both CST and TYPE must be scalars, or vectors
1905 of the same number of elements. If the value won't fit in the integer type,
1906 the results are undefined.</dd>
1908 <dt><b><tt>uitofp ( CST to TYPE )</tt></b></dt>
1909 <dd>Convert an unsigned integer constant to the corresponding floating point
1910 constant. TYPE must be a scalar or vector floating point type. CST must be of
1911 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1912 of the same number of elements. If the value won't fit in the floating point
1913 type, the results are undefined.</dd>
1915 <dt><b><tt>sitofp ( CST to TYPE )</tt></b></dt>
1916 <dd>Convert a signed integer constant to the corresponding floating point
1917 constant. TYPE must be a scalar or vector floating point type. CST must be of
1918 scalar or vector integer type. Both CST and TYPE must be scalars, or vectors
1919 of the same number of elements. If the value won't fit in the floating point
1920 type, the results are undefined.</dd>
1922 <dt><b><tt>ptrtoint ( CST to TYPE )</tt></b></dt>
1923 <dd>Convert a pointer typed constant to the corresponding integer constant
1924 TYPE must be an integer type. CST must be of pointer type. The CST value is
1925 zero extended, truncated, or unchanged to make it fit in TYPE.</dd>
1927 <dt><b><tt>inttoptr ( CST to TYPE )</tt></b></dt>
1928 <dd>Convert a integer constant to a pointer constant. TYPE must be a
1929 pointer type. CST must be of integer type. The CST value is zero extended,
1930 truncated, or unchanged to make it fit in a pointer size. This one is
1931 <i>really</i> dangerous!</dd>
1933 <dt><b><tt>bitcast ( CST to TYPE )</tt></b></dt>
1934 <dd>Convert a constant, CST, to another TYPE. The size of CST and TYPE must be
1935 identical (same number of bits). The conversion is done as if the CST value
1936 was stored to memory and read back as TYPE. In other words, no bits change
1937 with this operator, just the type. This can be used for conversion of
1938 vector types to any other type, as long as they have the same bit width. For
1939 pointers it is only valid to cast to another pointer type. It is not valid
1940 to bitcast to or from an aggregate type.
1943 <dt><b><tt>getelementptr ( CSTPTR, IDX0, IDX1, ... )</tt></b></dt>
1945 <dd>Perform the <a href="#i_getelementptr">getelementptr operation</a> on
1946 constants. As with the <a href="#i_getelementptr">getelementptr</a>
1947 instruction, the index list may have zero or more indexes, which are required
1948 to make sense for the type of "CSTPTR".</dd>
1950 <dt><b><tt>select ( COND, VAL1, VAL2 )</tt></b></dt>
1952 <dd>Perform the <a href="#i_select">select operation</a> on
1955 <dt><b><tt>icmp COND ( VAL1, VAL2 )</tt></b></dt>
1956 <dd>Performs the <a href="#i_icmp">icmp operation</a> on constants.</dd>
1958 <dt><b><tt>fcmp COND ( VAL1, VAL2 )</tt></b></dt>
1959 <dd>Performs the <a href="#i_fcmp">fcmp operation</a> on constants.</dd>
1961 <dt><b><tt>vicmp COND ( VAL1, VAL2 )</tt></b></dt>
1962 <dd>Performs the <a href="#i_vicmp">vicmp operation</a> on constants.</dd>
1964 <dt><b><tt>vfcmp COND ( VAL1, VAL2 )</tt></b></dt>
1965 <dd>Performs the <a href="#i_vfcmp">vfcmp operation</a> on constants.</dd>
1967 <dt><b><tt>extractelement ( VAL, IDX )</tt></b></dt>
1969 <dd>Perform the <a href="#i_extractelement">extractelement
1970 operation</a> on constants.</dd>
1972 <dt><b><tt>insertelement ( VAL, ELT, IDX )</tt></b></dt>
1974 <dd>Perform the <a href="#i_insertelement">insertelement
1975 operation</a> on constants.</dd>
1978 <dt><b><tt>shufflevector ( VEC1, VEC2, IDXMASK )</tt></b></dt>
1980 <dd>Perform the <a href="#i_shufflevector">shufflevector
1981 operation</a> on constants.</dd>
1983 <dt><b><tt>OPCODE ( LHS, RHS )</tt></b></dt>
1985 <dd>Perform the specified operation of the LHS and RHS constants. OPCODE may
1986 be any of the <a href="#binaryops">binary</a> or <a href="#bitwiseops">bitwise
1987 binary</a> operations. The constraints on operands are the same as those for
1988 the corresponding instruction (e.g. no bitwise operations on floating point
1989 values are allowed).</dd>
1993 <!-- *********************************************************************** -->
1994 <div class="doc_section"> <a name="othervalues">Other Values</a> </div>
1995 <!-- *********************************************************************** -->
1997 <!-- ======================================================================= -->
1998 <div class="doc_subsection">
1999 <a name="inlineasm">Inline Assembler Expressions</a>
2002 <div class="doc_text">
2005 LLVM supports inline assembler expressions (as opposed to <a href="#moduleasm">
2006 Module-Level Inline Assembly</a>) through the use of a special value. This
2007 value represents the inline assembler as a string (containing the instructions
2008 to emit), a list of operand constraints (stored as a string), and a flag that
2009 indicates whether or not the inline asm expression has side effects. An example
2010 inline assembler expression is:
2013 <div class="doc_code">
2015 i32 (i32) asm "bswap $0", "=r,r"
2020 Inline assembler expressions may <b>only</b> be used as the callee operand of
2021 a <a href="#i_call"><tt>call</tt> instruction</a>. Thus, typically we have:
2024 <div class="doc_code">
2026 %X = call i32 asm "<a href="#int_bswap">bswap</a> $0", "=r,r"(i32 %Y)
2031 Inline asms with side effects not visible in the constraint list must be marked
2032 as having side effects. This is done through the use of the
2033 '<tt>sideeffect</tt>' keyword, like so:
2036 <div class="doc_code">
2038 call void asm sideeffect "eieio", ""()
2042 <p>TODO: The format of the asm and constraints string still need to be
2043 documented here. Constraints on what can be done (e.g. duplication, moving, etc
2044 need to be documented). This is probably best done by reference to another
2045 document that covers inline asm from a holistic perspective.
2050 <!-- *********************************************************************** -->
2051 <div class="doc_section"> <a name="instref">Instruction Reference</a> </div>
2052 <!-- *********************************************************************** -->
2054 <div class="doc_text">
2056 <p>The LLVM instruction set consists of several different
2057 classifications of instructions: <a href="#terminators">terminator
2058 instructions</a>, <a href="#binaryops">binary instructions</a>,
2059 <a href="#bitwiseops">bitwise binary instructions</a>, <a
2060 href="#memoryops">memory instructions</a>, and <a href="#otherops">other
2061 instructions</a>.</p>
2065 <!-- ======================================================================= -->
2066 <div class="doc_subsection"> <a name="terminators">Terminator
2067 Instructions</a> </div>
2069 <div class="doc_text">
2071 <p>As mentioned <a href="#functionstructure">previously</a>, every
2072 basic block in a program ends with a "Terminator" instruction, which
2073 indicates which block should be executed after the current block is
2074 finished. These terminator instructions typically yield a '<tt>void</tt>'
2075 value: they produce control flow, not values (the one exception being
2076 the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction).</p>
2077 <p>There are six different terminator instructions: the '<a
2078 href="#i_ret"><tt>ret</tt></a>' instruction, the '<a href="#i_br"><tt>br</tt></a>'
2079 instruction, the '<a href="#i_switch"><tt>switch</tt></a>' instruction,
2080 the '<a href="#i_invoke"><tt>invoke</tt></a>' instruction, the '<a
2081 href="#i_unwind"><tt>unwind</tt></a>' instruction, and the '<a
2082 href="#i_unreachable"><tt>unreachable</tt></a>' instruction.</p>
2086 <!-- _______________________________________________________________________ -->
2087 <div class="doc_subsubsection"> <a name="i_ret">'<tt>ret</tt>'
2088 Instruction</a> </div>
2089 <div class="doc_text">
2092 ret <type> <value> <i>; Return a value from a non-void function</i>
2093 ret void <i>; Return from void function</i>
2098 <p>The '<tt>ret</tt>' instruction is used to return control flow (and
2099 optionally a value) from a function back to the caller.</p>
2100 <p>There are two forms of the '<tt>ret</tt>' instruction: one that
2101 returns a value and then causes control flow, and one that just causes
2102 control flow to occur.</p>
2106 <p>The '<tt>ret</tt>' instruction optionally accepts a single argument,
2107 the return value. The type of the return value must be a
2108 '<a href="#t_firstclass">first class</a>' type.</p>
2110 <p>A function is not <a href="#wellformed">well formed</a> if
2111 it it has a non-void return type and contains a '<tt>ret</tt>'
2112 instruction with no return value or a return value with a type that
2113 does not match its type, or if it has a void return type and contains
2114 a '<tt>ret</tt>' instruction with a return value.</p>
2118 <p>When the '<tt>ret</tt>' instruction is executed, control flow
2119 returns back to the calling function's context. If the caller is a "<a
2120 href="#i_call"><tt>call</tt></a>" instruction, execution continues at
2121 the instruction after the call. If the caller was an "<a
2122 href="#i_invoke"><tt>invoke</tt></a>" instruction, execution continues
2123 at the beginning of the "normal" destination block. If the instruction
2124 returns a value, that value shall set the call or invoke instruction's
2130 ret i32 5 <i>; Return an integer value of 5</i>
2131 ret void <i>; Return from a void function</i>
2132 ret { i32, i8 } { i32 4, i8 2 } <i>; Return an aggregate of values 4 and 2</i>
2135 <p>Note that the code generator does not yet fully support large
2136 return values. The specific sizes that are currently supported are
2137 dependent on the target. For integers, on 32-bit targets the limit
2138 is often 64 bits, and on 64-bit targets the limit is often 128 bits.
2139 For aggregate types, the current limits are dependent on the element
2140 types; for example targets are often limited to 2 total integer
2141 elements and 2 total floating-point elements.</p>
2144 <!-- _______________________________________________________________________ -->
2145 <div class="doc_subsubsection"> <a name="i_br">'<tt>br</tt>' Instruction</a> </div>
2146 <div class="doc_text">
2148 <pre> br i1 <cond>, label <iftrue>, label <iffalse><br> br label <dest> <i>; Unconditional branch</i>
2151 <p>The '<tt>br</tt>' instruction is used to cause control flow to
2152 transfer to a different basic block in the current function. There are
2153 two forms of this instruction, corresponding to a conditional branch
2154 and an unconditional branch.</p>
2156 <p>The conditional branch form of the '<tt>br</tt>' instruction takes a
2157 single '<tt>i1</tt>' value and two '<tt>label</tt>' values. The
2158 unconditional form of the '<tt>br</tt>' instruction takes a single
2159 '<tt>label</tt>' value as a target.</p>
2161 <p>Upon execution of a conditional '<tt>br</tt>' instruction, the '<tt>i1</tt>'
2162 argument is evaluated. If the value is <tt>true</tt>, control flows
2163 to the '<tt>iftrue</tt>' <tt>label</tt> argument. If "cond" is <tt>false</tt>,
2164 control flows to the '<tt>iffalse</tt>' <tt>label</tt> argument.</p>
2166 <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
2167 href="#i_ret">ret</a> i32 1<br>IfUnequal:<br> <a href="#i_ret">ret</a> i32 0<br></pre>
2169 <!-- _______________________________________________________________________ -->
2170 <div class="doc_subsubsection">
2171 <a name="i_switch">'<tt>switch</tt>' Instruction</a>
2174 <div class="doc_text">
2178 switch <intty> <value>, label <defaultdest> [ <intty> <val>, label <dest> ... ]
2183 <p>The '<tt>switch</tt>' instruction is used to transfer control flow to one of
2184 several different places. It is a generalization of the '<tt>br</tt>'
2185 instruction, allowing a branch to occur to one of many possible
2191 <p>The '<tt>switch</tt>' instruction uses three parameters: an integer
2192 comparison value '<tt>value</tt>', a default '<tt>label</tt>' destination, and
2193 an array of pairs of comparison value constants and '<tt>label</tt>'s. The
2194 table is not allowed to contain duplicate constant entries.</p>
2198 <p>The <tt>switch</tt> instruction specifies a table of values and
2199 destinations. When the '<tt>switch</tt>' instruction is executed, this
2200 table is searched for the given value. If the value is found, control flow is
2201 transfered to the corresponding destination; otherwise, control flow is
2202 transfered to the default destination.</p>
2204 <h5>Implementation:</h5>
2206 <p>Depending on properties of the target machine and the particular
2207 <tt>switch</tt> instruction, this instruction may be code generated in different
2208 ways. For example, it could be generated as a series of chained conditional
2209 branches or with a lookup table.</p>
2214 <i>; Emulate a conditional br instruction</i>
2215 %Val = <a href="#i_zext">zext</a> i1 %value to i32
2216 switch i32 %Val, label %truedest [ i32 0, label %falsedest ]
2218 <i>; Emulate an unconditional br instruction</i>
2219 switch i32 0, label %dest [ ]
2221 <i>; Implement a jump table:</i>
2222 switch i32 %val, label %otherwise [ i32 0, label %onzero
2224 i32 2, label %ontwo ]
2228 <!-- _______________________________________________________________________ -->
2229 <div class="doc_subsubsection">
2230 <a name="i_invoke">'<tt>invoke</tt>' Instruction</a>
2233 <div class="doc_text">
2238 <result> = invoke [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] <ptr to function ty> <function ptr val>(<function args>) [<a href="#fnattrs">fn attrs</a>]
2239 to label <normal label> unwind label <exception label>
2244 <p>The '<tt>invoke</tt>' instruction causes control to transfer to a specified
2245 function, with the possibility of control flow transfer to either the
2246 '<tt>normal</tt>' label or the
2247 '<tt>exception</tt>' label. If the callee function returns with the
2248 "<tt><a href="#i_ret">ret</a></tt>" instruction, control flow will return to the
2249 "normal" label. If the callee (or any indirect callees) returns with the "<a
2250 href="#i_unwind"><tt>unwind</tt></a>" instruction, control is interrupted and
2251 continued at the dynamically nearest "exception" label.</p>
2255 <p>This instruction requires several arguments:</p>
2259 The optional "cconv" marker indicates which <a href="#callingconv">calling
2260 convention</a> the call should use. If none is specified, the call defaults
2261 to using C calling conventions.
2264 <li>The optional <a href="#paramattrs">Parameter Attributes</a> list for
2265 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>',
2266 and '<tt>inreg</tt>' attributes are valid here.</li>
2268 <li>'<tt>ptr to function ty</tt>': shall be the signature of the pointer to
2269 function value being invoked. In most cases, this is a direct function
2270 invocation, but indirect <tt>invoke</tt>s are just as possible, branching off
2271 an arbitrary pointer to function value.
2274 <li>'<tt>function ptr val</tt>': An LLVM value containing a pointer to a
2275 function to be invoked. </li>
2277 <li>'<tt>function args</tt>': argument list whose types match the function
2278 signature argument types. If the function signature indicates the function
2279 accepts a variable number of arguments, the extra arguments can be
2282 <li>'<tt>normal label</tt>': the label reached when the called function
2283 executes a '<tt><a href="#i_ret">ret</a></tt>' instruction. </li>
2285 <li>'<tt>exception label</tt>': the label reached when a callee returns with
2286 the <a href="#i_unwind"><tt>unwind</tt></a> instruction. </li>
2288 <li>The optional <a href="#fnattrs">function attributes</a> list. Only
2289 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
2290 '<tt>readnone</tt>' attributes are valid here.</li>
2295 <p>This instruction is designed to operate as a standard '<tt><a
2296 href="#i_call">call</a></tt>' instruction in most regards. The primary
2297 difference is that it establishes an association with a label, which is used by
2298 the runtime library to unwind the stack.</p>
2300 <p>This instruction is used in languages with destructors to ensure that proper
2301 cleanup is performed in the case of either a <tt>longjmp</tt> or a thrown
2302 exception. Additionally, this is important for implementation of
2303 '<tt>catch</tt>' clauses in high-level languages that support them.</p>
2307 %retval = invoke i32 @Test(i32 15) to label %Continue
2308 unwind label %TestCleanup <i>; {i32}:retval set</i>
2309 %retval = invoke <a href="#callingconv">coldcc</a> i32 %Testfnptr(i32 15) to label %Continue
2310 unwind label %TestCleanup <i>; {i32}:retval set</i>
2315 <!-- _______________________________________________________________________ -->
2317 <div class="doc_subsubsection"> <a name="i_unwind">'<tt>unwind</tt>'
2318 Instruction</a> </div>
2320 <div class="doc_text">
2329 <p>The '<tt>unwind</tt>' instruction unwinds the stack, continuing control flow
2330 at the first callee in the dynamic call stack which used an <a
2331 href="#i_invoke"><tt>invoke</tt></a> instruction to perform the call. This is
2332 primarily used to implement exception handling.</p>
2336 <p>The '<tt>unwind</tt>' instruction causes execution of the current function to
2337 immediately halt. The dynamic call stack is then searched for the first <a
2338 href="#i_invoke"><tt>invoke</tt></a> instruction on the call stack. Once found,
2339 execution continues at the "exceptional" destination block specified by the
2340 <tt>invoke</tt> instruction. If there is no <tt>invoke</tt> instruction in the
2341 dynamic call chain, undefined behavior results.</p>
2344 <!-- _______________________________________________________________________ -->
2346 <div class="doc_subsubsection"> <a name="i_unreachable">'<tt>unreachable</tt>'
2347 Instruction</a> </div>
2349 <div class="doc_text">
2358 <p>The '<tt>unreachable</tt>' instruction has no defined semantics. This
2359 instruction is used to inform the optimizer that a particular portion of the
2360 code is not reachable. This can be used to indicate that the code after a
2361 no-return function cannot be reached, and other facts.</p>
2365 <p>The '<tt>unreachable</tt>' instruction has no defined semantics.</p>
2370 <!-- ======================================================================= -->
2371 <div class="doc_subsection"> <a name="binaryops">Binary Operations</a> </div>
2372 <div class="doc_text">
2373 <p>Binary operators are used to do most of the computation in a
2374 program. They require two operands of the same type, execute an operation on them, and
2375 produce a single value. The operands might represent
2376 multiple data, as is the case with the <a href="#t_vector">vector</a> data type.
2377 The result value has the same type as its operands.</p>
2378 <p>There are several different binary operators:</p>
2380 <!-- _______________________________________________________________________ -->
2381 <div class="doc_subsubsection">
2382 <a name="i_add">'<tt>add</tt>' Instruction</a>
2385 <div class="doc_text">
2390 <result> = add <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2395 <p>The '<tt>add</tt>' instruction returns the sum of its two operands.</p>
2399 <p>The two arguments to the '<tt>add</tt>' instruction must be <a
2400 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>, or
2401 <a href="#t_vector">vector</a> values. Both arguments must have identical
2406 <p>The value produced is the integer or floating point sum of the two
2409 <p>If an integer sum has unsigned overflow, the result returned is the
2410 mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2413 <p>Because LLVM integers use a two's complement representation, this
2414 instruction is appropriate for both signed and unsigned integers.</p>
2419 <result> = add i32 4, %var <i>; yields {i32}:result = 4 + %var</i>
2422 <!-- _______________________________________________________________________ -->
2423 <div class="doc_subsubsection">
2424 <a name="i_sub">'<tt>sub</tt>' Instruction</a>
2427 <div class="doc_text">
2432 <result> = sub <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2437 <p>The '<tt>sub</tt>' instruction returns the difference of its two
2440 <p>Note that the '<tt>sub</tt>' instruction is used to represent the
2441 '<tt>neg</tt>' instruction present in most other intermediate
2442 representations.</p>
2446 <p>The two arguments to the '<tt>sub</tt>' instruction must be <a
2447 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2448 or <a href="#t_vector">vector</a> values. Both arguments must have identical
2453 <p>The value produced is the integer or floating point difference of
2454 the two operands.</p>
2456 <p>If an integer difference has unsigned overflow, the result returned is the
2457 mathematical result modulo 2<sup>n</sup>, where n is the bit width of
2460 <p>Because LLVM integers use a two's complement representation, this
2461 instruction is appropriate for both signed and unsigned integers.</p>
2465 <result> = sub i32 4, %var <i>; yields {i32}:result = 4 - %var</i>
2466 <result> = sub i32 0, %val <i>; yields {i32}:result = -%var</i>
2470 <!-- _______________________________________________________________________ -->
2471 <div class="doc_subsubsection">
2472 <a name="i_mul">'<tt>mul</tt>' Instruction</a>
2475 <div class="doc_text">
2478 <pre> <result> = mul <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2481 <p>The '<tt>mul</tt>' instruction returns the product of its two
2486 <p>The two arguments to the '<tt>mul</tt>' instruction must be <a
2487 href="#t_integer">integer</a>, <a href="#t_floating">floating point</a>,
2488 or <a href="#t_vector">vector</a> values. Both arguments must have identical
2493 <p>The value produced is the integer or floating point product of the
2496 <p>If the result of an integer multiplication has unsigned overflow,
2497 the result returned is the mathematical result modulo
2498 2<sup>n</sup>, where n is the bit width of the result.</p>
2499 <p>Because LLVM integers use a two's complement representation, and the
2500 result is the same width as the operands, this instruction returns the
2501 correct result for both signed and unsigned integers. If a full product
2502 (e.g. <tt>i32</tt>x<tt>i32</tt>-><tt>i64</tt>) is needed, the operands
2503 should be sign-extended or zero-extended as appropriate to the
2504 width of the full product.</p>
2506 <pre> <result> = mul i32 4, %var <i>; yields {i32}:result = 4 * %var</i>
2510 <!-- _______________________________________________________________________ -->
2511 <div class="doc_subsubsection"> <a name="i_udiv">'<tt>udiv</tt>' Instruction
2513 <div class="doc_text">
2515 <pre> <result> = udiv <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2518 <p>The '<tt>udiv</tt>' instruction returns the quotient of its two
2523 <p>The two arguments to the '<tt>udiv</tt>' instruction must be
2524 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2525 values. Both arguments must have identical types.</p>
2529 <p>The value produced is the unsigned integer quotient of the two operands.</p>
2530 <p>Note that unsigned integer division and signed integer division are distinct
2531 operations; for signed integer division, use '<tt>sdiv</tt>'.</p>
2532 <p>Division by zero leads to undefined behavior.</p>
2534 <pre> <result> = udiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
2537 <!-- _______________________________________________________________________ -->
2538 <div class="doc_subsubsection"> <a name="i_sdiv">'<tt>sdiv</tt>' Instruction
2540 <div class="doc_text">
2543 <result> = sdiv <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2548 <p>The '<tt>sdiv</tt>' instruction returns the quotient of its two
2553 <p>The two arguments to the '<tt>sdiv</tt>' instruction must be
2554 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2555 values. Both arguments must have identical types.</p>
2558 <p>The value produced is the signed integer quotient of the two operands rounded towards zero.</p>
2559 <p>Note that signed integer division and unsigned integer division are distinct
2560 operations; for unsigned integer division, use '<tt>udiv</tt>'.</p>
2561 <p>Division by zero leads to undefined behavior. Overflow also leads to
2562 undefined behavior; this is a rare case, but can occur, for example,
2563 by doing a 32-bit division of -2147483648 by -1.</p>
2565 <pre> <result> = sdiv i32 4, %var <i>; yields {i32}:result = 4 / %var</i>
2568 <!-- _______________________________________________________________________ -->
2569 <div class="doc_subsubsection"> <a name="i_fdiv">'<tt>fdiv</tt>'
2570 Instruction</a> </div>
2571 <div class="doc_text">
2574 <result> = fdiv <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2578 <p>The '<tt>fdiv</tt>' instruction returns the quotient of its two
2583 <p>The two arguments to the '<tt>fdiv</tt>' instruction must be
2584 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2585 of floating point values. Both arguments must have identical types.</p>
2589 <p>The value produced is the floating point quotient of the two operands.</p>
2594 <result> = fdiv float 4.0, %var <i>; yields {float}:result = 4.0 / %var</i>
2598 <!-- _______________________________________________________________________ -->
2599 <div class="doc_subsubsection"> <a name="i_urem">'<tt>urem</tt>' Instruction</a>
2601 <div class="doc_text">
2603 <pre> <result> = urem <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2606 <p>The '<tt>urem</tt>' instruction returns the remainder from the
2607 unsigned division of its two arguments.</p>
2609 <p>The two arguments to the '<tt>urem</tt>' instruction must be
2610 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2611 values. Both arguments must have identical types.</p>
2613 <p>This instruction returns the unsigned integer <i>remainder</i> of a division.
2614 This instruction always performs an unsigned division to get the remainder.</p>
2615 <p>Note that unsigned integer remainder and signed integer remainder are
2616 distinct operations; for signed integer remainder, use '<tt>srem</tt>'.</p>
2617 <p>Taking the remainder of a division by zero leads to undefined behavior.</p>
2619 <pre> <result> = urem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2623 <!-- _______________________________________________________________________ -->
2624 <div class="doc_subsubsection">
2625 <a name="i_srem">'<tt>srem</tt>' Instruction</a>
2628 <div class="doc_text">
2633 <result> = srem <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2638 <p>The '<tt>srem</tt>' instruction returns the remainder from the
2639 signed division of its two operands. This instruction can also take
2640 <a href="#t_vector">vector</a> versions of the values in which case
2641 the elements must be integers.</p>
2645 <p>The two arguments to the '<tt>srem</tt>' instruction must be
2646 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2647 values. Both arguments must have identical types.</p>
2651 <p>This instruction returns the <i>remainder</i> of a division (where the result
2652 has the same sign as the dividend, <tt>op1</tt>), not the <i>modulo</i>
2653 operator (where the result has the same sign as the divisor, <tt>op2</tt>) of
2654 a value. For more information about the difference, see <a
2655 href="http://mathforum.org/dr.math/problems/anne.4.28.99.html">The
2656 Math Forum</a>. For a table of how this is implemented in various languages,
2657 please see <a href="http://en.wikipedia.org/wiki/Modulo_operation">
2658 Wikipedia: modulo operation</a>.</p>
2659 <p>Note that signed integer remainder and unsigned integer remainder are
2660 distinct operations; for unsigned integer remainder, use '<tt>urem</tt>'.</p>
2661 <p>Taking the remainder of a division by zero leads to undefined behavior.
2662 Overflow also leads to undefined behavior; this is a rare case, but can occur,
2663 for example, by taking the remainder of a 32-bit division of -2147483648 by -1.
2664 (The remainder doesn't actually overflow, but this rule lets srem be
2665 implemented using instructions that return both the result of the division
2666 and the remainder.)</p>
2668 <pre> <result> = srem i32 4, %var <i>; yields {i32}:result = 4 % %var</i>
2672 <!-- _______________________________________________________________________ -->
2673 <div class="doc_subsubsection">
2674 <a name="i_frem">'<tt>frem</tt>' Instruction</a> </div>
2676 <div class="doc_text">
2679 <pre> <result> = frem <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2682 <p>The '<tt>frem</tt>' instruction returns the remainder from the
2683 division of its two operands.</p>
2685 <p>The two arguments to the '<tt>frem</tt>' instruction must be
2686 <a href="#t_floating">floating point</a> or <a href="#t_vector">vector</a>
2687 of floating point values. Both arguments must have identical types.</p>
2691 <p>This instruction returns the <i>remainder</i> of a division.
2692 The remainder has the same sign as the dividend.</p>
2697 <result> = frem float 4.0, %var <i>; yields {float}:result = 4.0 % %var</i>
2701 <!-- ======================================================================= -->
2702 <div class="doc_subsection"> <a name="bitwiseops">Bitwise Binary
2703 Operations</a> </div>
2704 <div class="doc_text">
2705 <p>Bitwise binary operators are used to do various forms of
2706 bit-twiddling in a program. They are generally very efficient
2707 instructions and can commonly be strength reduced from other
2708 instructions. They require two operands of the same type, execute an operation on them,
2709 and produce a single value. The resulting value is the same type as its operands.</p>
2712 <!-- _______________________________________________________________________ -->
2713 <div class="doc_subsubsection"> <a name="i_shl">'<tt>shl</tt>'
2714 Instruction</a> </div>
2715 <div class="doc_text">
2717 <pre> <result> = shl <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2722 <p>The '<tt>shl</tt>' instruction returns the first operand shifted to
2723 the left a specified number of bits.</p>
2727 <p>Both arguments to the '<tt>shl</tt>' instruction must be the same <a
2728 href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2729 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
2733 <p>The value produced is <tt>op1</tt> * 2<sup><tt>op2</tt></sup> mod 2<sup>n</sup>,
2734 where n is the width of the result. If <tt>op2</tt> is (statically or dynamically) negative or
2735 equal to or larger than the number of bits in <tt>op1</tt>, the result is undefined.
2736 If the arguments are vectors, each vector element of <tt>op1</tt> is shifted by the
2737 corresponding shift amount in <tt>op2</tt>.</p>
2739 <h5>Example:</h5><pre>
2740 <result> = shl i32 4, %var <i>; yields {i32}: 4 << %var</i>
2741 <result> = shl i32 4, 2 <i>; yields {i32}: 16</i>
2742 <result> = shl i32 1, 10 <i>; yields {i32}: 1024</i>
2743 <result> = shl i32 1, 32 <i>; undefined</i>
2744 <result> = shl <2 x i32> < i32 1, i32 1>, < i32 1, i32 2> <i>; yields: result=<2 x i32> < i32 2, i32 4></i>
2747 <!-- _______________________________________________________________________ -->
2748 <div class="doc_subsubsection"> <a name="i_lshr">'<tt>lshr</tt>'
2749 Instruction</a> </div>
2750 <div class="doc_text">
2752 <pre> <result> = lshr <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2756 <p>The '<tt>lshr</tt>' instruction (logical shift right) returns the first
2757 operand shifted to the right a specified number of bits with zero fill.</p>
2760 <p>Both arguments to the '<tt>lshr</tt>' instruction must be the same
2761 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2762 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
2766 <p>This instruction always performs a logical shift right operation. The most
2767 significant bits of the result will be filled with zero bits after the
2768 shift. If <tt>op2</tt> is (statically or dynamically) equal to or larger than
2769 the number of bits in <tt>op1</tt>, the result is undefined. If the arguments are
2770 vectors, each vector element of <tt>op1</tt> is shifted by the corresponding shift
2771 amount in <tt>op2</tt>.</p>
2775 <result> = lshr i32 4, 1 <i>; yields {i32}:result = 2</i>
2776 <result> = lshr i32 4, 2 <i>; yields {i32}:result = 1</i>
2777 <result> = lshr i8 4, 3 <i>; yields {i8}:result = 0</i>
2778 <result> = lshr i8 -2, 1 <i>; yields {i8}:result = 0x7FFFFFFF </i>
2779 <result> = lshr i32 1, 32 <i>; undefined</i>
2780 <result> = lshr <2 x i32> < i32 -2, i32 4>, < i32 1, i32 2> <i>; yields: result=<2 x i32> < i32 0x7FFFFFFF, i32 1></i>
2784 <!-- _______________________________________________________________________ -->
2785 <div class="doc_subsubsection"> <a name="i_ashr">'<tt>ashr</tt>'
2786 Instruction</a> </div>
2787 <div class="doc_text">
2790 <pre> <result> = ashr <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2794 <p>The '<tt>ashr</tt>' instruction (arithmetic shift right) returns the first
2795 operand shifted to the right a specified number of bits with sign extension.</p>
2798 <p>Both arguments to the '<tt>ashr</tt>' instruction must be the same
2799 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2800 type. '<tt>op2</tt>' is treated as an unsigned value.</p>
2803 <p>This instruction always performs an arithmetic shift right operation,
2804 The most significant bits of the result will be filled with the sign bit
2805 of <tt>op1</tt>. If <tt>op2</tt> is (statically or dynamically) equal to or
2806 larger than the number of bits in <tt>op1</tt>, the result is undefined. If the
2807 arguments are vectors, each vector element of <tt>op1</tt> is shifted by the
2808 corresponding shift amount in <tt>op2</tt>.</p>
2812 <result> = ashr i32 4, 1 <i>; yields {i32}:result = 2</i>
2813 <result> = ashr i32 4, 2 <i>; yields {i32}:result = 1</i>
2814 <result> = ashr i8 4, 3 <i>; yields {i8}:result = 0</i>
2815 <result> = ashr i8 -2, 1 <i>; yields {i8}:result = -1</i>
2816 <result> = ashr i32 1, 32 <i>; undefined</i>
2817 <result> = ashr <2 x i32> < i32 -2, i32 4>, < i32 1, i32 3> <i>; yields: result=<2 x i32> < i32 -1, i32 0></i>
2821 <!-- _______________________________________________________________________ -->
2822 <div class="doc_subsubsection"> <a name="i_and">'<tt>and</tt>'
2823 Instruction</a> </div>
2825 <div class="doc_text">
2830 <result> = and <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2835 <p>The '<tt>and</tt>' instruction returns the bitwise logical and of
2836 its two operands.</p>
2840 <p>The two arguments to the '<tt>and</tt>' instruction must be
2841 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2842 values. Both arguments must have identical types.</p>
2845 <p>The truth table used for the '<tt>and</tt>' instruction is:</p>
2848 <table border="1" cellspacing="0" cellpadding="4">
2880 <result> = and i32 4, %var <i>; yields {i32}:result = 4 & %var</i>
2881 <result> = and i32 15, 40 <i>; yields {i32}:result = 8</i>
2882 <result> = and i32 4, 8 <i>; yields {i32}:result = 0</i>
2885 <!-- _______________________________________________________________________ -->
2886 <div class="doc_subsubsection"> <a name="i_or">'<tt>or</tt>' Instruction</a> </div>
2887 <div class="doc_text">
2889 <pre> <result> = or <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2892 <p>The '<tt>or</tt>' instruction returns the bitwise logical inclusive
2893 or of its two operands.</p>
2896 <p>The two arguments to the '<tt>or</tt>' instruction must be
2897 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2898 values. Both arguments must have identical types.</p>
2900 <p>The truth table used for the '<tt>or</tt>' instruction is:</p>
2903 <table border="1" cellspacing="0" cellpadding="4">
2934 <pre> <result> = or i32 4, %var <i>; yields {i32}:result = 4 | %var</i>
2935 <result> = or i32 15, 40 <i>; yields {i32}:result = 47</i>
2936 <result> = or i32 4, 8 <i>; yields {i32}:result = 12</i>
2939 <!-- _______________________________________________________________________ -->
2940 <div class="doc_subsubsection"> <a name="i_xor">'<tt>xor</tt>'
2941 Instruction</a> </div>
2942 <div class="doc_text">
2944 <pre> <result> = xor <ty> <op1>, <op2> <i>; yields {ty}:result</i>
2947 <p>The '<tt>xor</tt>' instruction returns the bitwise logical exclusive
2948 or of its two operands. The <tt>xor</tt> is used to implement the
2949 "one's complement" operation, which is the "~" operator in C.</p>
2951 <p>The two arguments to the '<tt>xor</tt>' instruction must be
2952 <a href="#t_integer">integer</a> or <a href="#t_vector">vector</a> of integer
2953 values. Both arguments must have identical types.</p>
2957 <p>The truth table used for the '<tt>xor</tt>' instruction is:</p>
2960 <table border="1" cellspacing="0" cellpadding="4">
2992 <pre> <result> = xor i32 4, %var <i>; yields {i32}:result = 4 ^ %var</i>
2993 <result> = xor i32 15, 40 <i>; yields {i32}:result = 39</i>
2994 <result> = xor i32 4, 8 <i>; yields {i32}:result = 12</i>
2995 <result> = xor i32 %V, -1 <i>; yields {i32}:result = ~%V</i>
2999 <!-- ======================================================================= -->
3000 <div class="doc_subsection">
3001 <a name="vectorops">Vector Operations</a>
3004 <div class="doc_text">
3006 <p>LLVM supports several instructions to represent vector operations in a
3007 target-independent manner. These instructions cover the element-access and
3008 vector-specific operations needed to process vectors effectively. While LLVM
3009 does directly support these vector operations, many sophisticated algorithms
3010 will want to use target-specific intrinsics to take full advantage of a specific
3015 <!-- _______________________________________________________________________ -->
3016 <div class="doc_subsubsection">
3017 <a name="i_extractelement">'<tt>extractelement</tt>' Instruction</a>
3020 <div class="doc_text">
3025 <result> = extractelement <n x <ty>> <val>, i32 <idx> <i>; yields <ty></i>
3031 The '<tt>extractelement</tt>' instruction extracts a single scalar
3032 element from a vector at a specified index.
3039 The first operand of an '<tt>extractelement</tt>' instruction is a
3040 value of <a href="#t_vector">vector</a> type. The second operand is
3041 an index indicating the position from which to extract the element.
3042 The index may be a variable.</p>
3047 The result is a scalar of the same type as the element type of
3048 <tt>val</tt>. Its value is the value at position <tt>idx</tt> of
3049 <tt>val</tt>. If <tt>idx</tt> exceeds the length of <tt>val</tt>, the
3050 results are undefined.
3056 %result = extractelement <4 x i32> %vec, i32 0 <i>; yields i32</i>
3061 <!-- _______________________________________________________________________ -->
3062 <div class="doc_subsubsection">
3063 <a name="i_insertelement">'<tt>insertelement</tt>' Instruction</a>
3066 <div class="doc_text">
3071 <result> = insertelement <n x <ty>> <val>, <ty> <elt>, i32 <idx> <i>; yields <n x <ty>></i>
3077 The '<tt>insertelement</tt>' instruction inserts a scalar
3078 element into a vector at a specified index.
3085 The first operand of an '<tt>insertelement</tt>' instruction is a
3086 value of <a href="#t_vector">vector</a> type. The second operand is a
3087 scalar value whose type must equal the element type of the first
3088 operand. The third operand is an index indicating the position at
3089 which to insert the value. The index may be a variable.</p>
3094 The result is a vector of the same type as <tt>val</tt>. Its
3095 element values are those of <tt>val</tt> except at position
3096 <tt>idx</tt>, where it gets the value <tt>elt</tt>. If <tt>idx</tt>
3097 exceeds the length of <tt>val</tt>, the results are undefined.
3103 %result = insertelement <4 x i32> %vec, i32 1, i32 0 <i>; yields <4 x i32></i>
3107 <!-- _______________________________________________________________________ -->
3108 <div class="doc_subsubsection">
3109 <a name="i_shufflevector">'<tt>shufflevector</tt>' Instruction</a>
3112 <div class="doc_text">
3117 <result> = shufflevector <n x <ty>> <v1>, <n x <ty>> <v2>, <m x i32> <mask> <i>; yields <m x <ty>></i>
3123 The '<tt>shufflevector</tt>' instruction constructs a permutation of elements
3124 from two input vectors, returning a vector with the same element type as
3125 the input and length that is the same as the shuffle mask.
3131 The first two operands of a '<tt>shufflevector</tt>' instruction are vectors
3132 with types that match each other. The third argument is a shuffle mask whose
3133 element type is always 'i32'. The result of the instruction is a vector whose
3134 length is the same as the shuffle mask and whose element type is the same as
3135 the element type of the first two operands.
3139 The shuffle mask operand is required to be a constant vector with either
3140 constant integer or undef values.
3146 The elements of the two input vectors are numbered from left to right across
3147 both of the vectors. The shuffle mask operand specifies, for each element of
3148 the result vector, which element of the two input vectors the result element
3149 gets. The element selector may be undef (meaning "don't care") and the second
3150 operand may be undef if performing a shuffle from only one vector.
3156 %result = shufflevector <4 x i32> %v1, <4 x i32> %v2,
3157 <4 x i32> <i32 0, i32 4, i32 1, i32 5> <i>; yields <4 x i32></i>
3158 %result = shufflevector <4 x i32> %v1, <4 x i32> undef,
3159 <4 x i32> <i32 0, i32 1, i32 2, i32 3> <i>; yields <4 x i32></i> - Identity shuffle.
3160 %result = shufflevector <8 x i32> %v1, <8 x i32> undef,
3161 <4 x i32> <i32 0, i32 1, i32 2, i32 3> <i>; yields <4 x i32></i>
3162 %result = shufflevector <4 x i32> %v1, <4 x i32> %v2,
3163 <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7 > <i>; yields <8 x i32></i>
3168 <!-- ======================================================================= -->
3169 <div class="doc_subsection">
3170 <a name="aggregateops">Aggregate Operations</a>
3173 <div class="doc_text">
3175 <p>LLVM supports several instructions for working with aggregate values.
3180 <!-- _______________________________________________________________________ -->
3181 <div class="doc_subsubsection">
3182 <a name="i_extractvalue">'<tt>extractvalue</tt>' Instruction</a>
3185 <div class="doc_text">
3190 <result> = extractvalue <aggregate type> <val>, <idx>{, <idx>}*
3196 The '<tt>extractvalue</tt>' instruction extracts the value of a struct field
3197 or array element from an aggregate value.
3204 The first operand of an '<tt>extractvalue</tt>' instruction is a
3205 value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a>
3206 type. The operands are constant indices to specify which value to extract
3207 in a similar manner as indices in a
3208 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
3214 The result is the value at the position in the aggregate specified by
3221 %result = extractvalue {i32, float} %agg, 0 <i>; yields i32</i>
3226 <!-- _______________________________________________________________________ -->
3227 <div class="doc_subsubsection">
3228 <a name="i_insertvalue">'<tt>insertvalue</tt>' Instruction</a>
3231 <div class="doc_text">
3236 <result> = insertvalue <aggregate type> <val>, <ty> <val>, <idx> <i>; yields <n x <ty>></i>
3242 The '<tt>insertvalue</tt>' instruction inserts a value
3243 into a struct field or array element in an aggregate.
3250 The first operand of an '<tt>insertvalue</tt>' instruction is a
3251 value of <a href="#t_struct">struct</a> or <a href="#t_array">array</a> type.
3252 The second operand is a first-class value to insert.
3253 The following operands are constant indices
3254 indicating the position at which to insert the value in a similar manner as
3256 '<tt><a href="#i_getelementptr">getelementptr</a></tt>' instruction.
3257 The value to insert must have the same type as the value identified
3264 The result is an aggregate of the same type as <tt>val</tt>. Its
3265 value is that of <tt>val</tt> except that the value at the position
3266 specified by the indices is that of <tt>elt</tt>.
3272 %result = insertvalue {i32, float} %agg, i32 1, 0 <i>; yields {i32, float}</i>
3277 <!-- ======================================================================= -->
3278 <div class="doc_subsection">
3279 <a name="memoryops">Memory Access and Addressing Operations</a>
3282 <div class="doc_text">
3284 <p>A key design point of an SSA-based representation is how it
3285 represents memory. In LLVM, no memory locations are in SSA form, which
3286 makes things very simple. This section describes how to read, write,
3287 allocate, and free memory in LLVM.</p>
3291 <!-- _______________________________________________________________________ -->
3292 <div class="doc_subsubsection">
3293 <a name="i_malloc">'<tt>malloc</tt>' Instruction</a>
3296 <div class="doc_text">
3301 <result> = malloc <type>[, i32 <NumElements>][, align <alignment>] <i>; yields {type*}:result</i>
3306 <p>The '<tt>malloc</tt>' instruction allocates memory from the system
3307 heap and returns a pointer to it. The object is always allocated in the generic
3308 address space (address space zero).</p>
3312 <p>The '<tt>malloc</tt>' instruction allocates
3313 <tt>sizeof(<type>)*NumElements</tt>
3314 bytes of memory from the operating system and returns a pointer of the
3315 appropriate type to the program. If "NumElements" is specified, it is the
3316 number of elements allocated, otherwise "NumElements" is defaulted to be one.
3317 If a constant alignment is specified, the value result of the allocation is guaranteed to
3318 be aligned to at least that boundary. If not specified, or if zero, the target can
3319 choose to align the allocation on any convenient boundary.</p>
3321 <p>'<tt>type</tt>' must be a sized type.</p>
3325 <p>Memory is allocated using the system "<tt>malloc</tt>" function, and
3326 a pointer is returned. The result of a zero byte allocation is undefined. The
3327 result is null if there is insufficient memory available.</p>
3332 %array = malloc [4 x i8] <i>; yields {[%4 x i8]*}:array</i>
3334 %size = <a href="#i_add">add</a> i32 2, 2 <i>; yields {i32}:size = i32 4</i>
3335 %array1 = malloc i8, i32 4 <i>; yields {i8*}:array1</i>
3336 %array2 = malloc [12 x i8], i32 %size <i>; yields {[12 x i8]*}:array2</i>
3337 %array3 = malloc i32, i32 4, align 1024 <i>; yields {i32*}:array3</i>
3338 %array4 = malloc i32, align 1024 <i>; yields {i32*}:array4</i>
3341 <p>Note that the code generator does not yet respect the
3342 alignment value.</p>
3346 <!-- _______________________________________________________________________ -->
3347 <div class="doc_subsubsection">
3348 <a name="i_free">'<tt>free</tt>' Instruction</a>
3351 <div class="doc_text">
3356 free <type> <value> <i>; yields {void}</i>
3361 <p>The '<tt>free</tt>' instruction returns memory back to the unused
3362 memory heap to be reallocated in the future.</p>
3366 <p>'<tt>value</tt>' shall be a pointer value that points to a value
3367 that was allocated with the '<tt><a href="#i_malloc">malloc</a></tt>'
3372 <p>Access to the memory pointed to by the pointer is no longer defined
3373 after this instruction executes. If the pointer is null, the operation
3379 %array = <a href="#i_malloc">malloc</a> [4 x i8] <i>; yields {[4 x i8]*}:array</i>
3380 free [4 x i8]* %array
3384 <!-- _______________________________________________________________________ -->
3385 <div class="doc_subsubsection">
3386 <a name="i_alloca">'<tt>alloca</tt>' Instruction</a>
3389 <div class="doc_text">
3394 <result> = alloca <type>[, i32 <NumElements>][, align <alignment>] <i>; yields {type*}:result</i>
3399 <p>The '<tt>alloca</tt>' instruction allocates memory on the stack frame of the
3400 currently executing function, to be automatically released when this function
3401 returns to its caller. The object is always allocated in the generic address
3402 space (address space zero).</p>
3406 <p>The '<tt>alloca</tt>' instruction allocates <tt>sizeof(<type>)*NumElements</tt>
3407 bytes of memory on the runtime stack, returning a pointer of the
3408 appropriate type to the program. If "NumElements" is specified, it is the
3409 number of elements allocated, otherwise "NumElements" is defaulted to be one.
3410 If a constant alignment is specified, the value result of the allocation is guaranteed
3411 to be aligned to at least that boundary. If not specified, or if zero, the target
3412 can choose to align the allocation on any convenient boundary.</p>
3414 <p>'<tt>type</tt>' may be any sized type.</p>
3418 <p>Memory is allocated; a pointer is returned. The operation is undefiend if
3419 there is insufficient stack space for the allocation. '<tt>alloca</tt>'d
3420 memory is automatically released when the function returns. The '<tt>alloca</tt>'
3421 instruction is commonly used to represent automatic variables that must
3422 have an address available. When the function returns (either with the <tt><a
3423 href="#i_ret">ret</a></tt> or <tt><a href="#i_unwind">unwind</a></tt>
3424 instructions), the memory is reclaimed. Allocating zero bytes
3425 is legal, but the result is undefined.</p>
3430 %ptr = alloca i32 <i>; yields {i32*}:ptr</i>
3431 %ptr = alloca i32, i32 4 <i>; yields {i32*}:ptr</i>
3432 %ptr = alloca i32, i32 4, align 1024 <i>; yields {i32*}:ptr</i>
3433 %ptr = alloca i32, align 1024 <i>; yields {i32*}:ptr</i>
3437 <!-- _______________________________________________________________________ -->
3438 <div class="doc_subsubsection"> <a name="i_load">'<tt>load</tt>'
3439 Instruction</a> </div>
3440 <div class="doc_text">
3442 <pre> <result> = load <ty>* <pointer>[, align <alignment>]<br> <result> = volatile load <ty>* <pointer>[, align <alignment>]<br></pre>
3444 <p>The '<tt>load</tt>' instruction is used to read from memory.</p>
3446 <p>The argument to the '<tt>load</tt>' instruction specifies the memory
3447 address from which to load. The pointer must point to a <a
3448 href="#t_firstclass">first class</a> type. If the <tt>load</tt> is
3449 marked as <tt>volatile</tt>, then the optimizer is not allowed to modify
3450 the number or order of execution of this <tt>load</tt> with other
3451 volatile <tt>load</tt> and <tt><a href="#i_store">store</a></tt>
3454 The optional constant "align" argument specifies the alignment of the operation
3455 (that is, the alignment of the memory address). A value of 0 or an
3456 omitted "align" argument means that the operation has the preferential
3457 alignment for the target. It is the responsibility of the code emitter
3458 to ensure that the alignment information is correct. Overestimating
3459 the alignment results in an undefined behavior. Underestimating the
3460 alignment may produce less efficient code. An alignment of 1 is always
3464 <p>The location of memory pointed to is loaded.</p>
3466 <pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
3468 href="#i_store">store</a> i32 3, i32* %ptr <i>; yields {void}</i>
3469 %val = load i32* %ptr <i>; yields {i32}:val = i32 3</i>
3472 <!-- _______________________________________________________________________ -->
3473 <div class="doc_subsubsection"> <a name="i_store">'<tt>store</tt>'
3474 Instruction</a> </div>
3475 <div class="doc_text">
3477 <pre> store <ty> <value>, <ty>* <pointer>[, align <alignment>] <i>; yields {void}</i>
3478 volatile store <ty> <value>, <ty>* <pointer>[, align <alignment>] <i>; yields {void}</i>
3481 <p>The '<tt>store</tt>' instruction is used to write to memory.</p>
3483 <p>There are two arguments to the '<tt>store</tt>' instruction: a value
3484 to store and an address at which to store it. The type of the '<tt><pointer></tt>'
3485 operand must be a pointer to the <a href="#t_firstclass">first class</a> type
3486 of the '<tt><value></tt>'
3487 operand. If the <tt>store</tt> is marked as <tt>volatile</tt>, then the
3488 optimizer is not allowed to modify the number or order of execution of
3489 this <tt>store</tt> with other volatile <tt>load</tt> and <tt><a
3490 href="#i_store">store</a></tt> instructions.</p>
3492 The optional constant "align" argument specifies the alignment of the operation
3493 (that is, the alignment of the memory address). A value of 0 or an
3494 omitted "align" argument means that the operation has the preferential
3495 alignment for the target. It is the responsibility of the code emitter
3496 to ensure that the alignment information is correct. Overestimating
3497 the alignment results in an undefined behavior. Underestimating the
3498 alignment may produce less efficient code. An alignment of 1 is always
3502 <p>The contents of memory are updated to contain '<tt><value></tt>'
3503 at the location specified by the '<tt><pointer></tt>' operand.</p>
3505 <pre> %ptr = <a href="#i_alloca">alloca</a> i32 <i>; yields {i32*}:ptr</i>
3506 store i32 3, i32* %ptr <i>; yields {void}</i>
3507 %val = <a href="#i_load">load</a> i32* %ptr <i>; yields {i32}:val = i32 3</i>
3511 <!-- _______________________________________________________________________ -->
3512 <div class="doc_subsubsection">
3513 <a name="i_getelementptr">'<tt>getelementptr</tt>' Instruction</a>
3516 <div class="doc_text">
3519 <result> = getelementptr <pty>* <ptrval>{, <ty> <idx>}*
3525 The '<tt>getelementptr</tt>' instruction is used to get the address of a
3526 subelement of an aggregate data structure. It performs address calculation only
3527 and does not access memory.</p>
3531 <p>The first argument is always a pointer, and forms the basis of the
3532 calculation. The remaining arguments are indices, that indicate which of the
3533 elements of the aggregate object are indexed. The interpretation of each index
3534 is dependent on the type being indexed into. The first index always indexes the
3535 pointer value given as the first argument, the second index indexes a value of
3536 the type pointed to (not necessarily the value directly pointed to, since the
3537 first index can be non-zero), etc. The first type indexed into must be a pointer
3538 value, subsequent types can be arrays, vectors and structs. Note that subsequent
3539 types being indexed into can never be pointers, since that would require loading
3540 the pointer before continuing calculation.</p>
3542 <p>The type of each index argument depends on the type it is indexing into.
3543 When indexing into a (packed) structure, only <tt>i32</tt> integer
3544 <b>constants</b> are allowed. When indexing into an array, pointer or vector,
3545 only integers of 32 or 64 bits are allowed (also non-constants). 32-bit values
3546 will be sign extended to 64-bits if required.</p>
3548 <p>For example, let's consider a C code fragment and how it gets
3549 compiled to LLVM:</p>
3551 <div class="doc_code">
3564 int *foo(struct ST *s) {
3565 return &s[1].Z.B[5][13];
3570 <p>The LLVM code generated by the GCC frontend is:</p>
3572 <div class="doc_code">
3574 %RT = <a href="#namedtypes">type</a> { i8 , [10 x [20 x i32]], i8 }
3575 %ST = <a href="#namedtypes">type</a> { i32, double, %RT }
3577 define i32* %foo(%ST* %s) {
3579 %reg = getelementptr %ST* %s, i32 1, i32 2, i32 1, i32 5, i32 13
3587 <p>In the example above, the first index is indexing into the '<tt>%ST*</tt>'
3588 type, which is a pointer, yielding a '<tt>%ST</tt>' = '<tt>{ i32, double, %RT
3589 }</tt>' type, a structure. The second index indexes into the third element of
3590 the structure, yielding a '<tt>%RT</tt>' = '<tt>{ i8 , [10 x [20 x i32]],
3591 i8 }</tt>' type, another structure. The third index indexes into the second
3592 element of the structure, yielding a '<tt>[10 x [20 x i32]]</tt>' type, an
3593 array. The two dimensions of the array are subscripted into, yielding an
3594 '<tt>i32</tt>' type. The '<tt>getelementptr</tt>' instruction returns a pointer
3595 to this element, thus computing a value of '<tt>i32*</tt>' type.</p>
3597 <p>Note that it is perfectly legal to index partially through a
3598 structure, returning a pointer to an inner element. Because of this,
3599 the LLVM code for the given testcase is equivalent to:</p>
3602 define i32* %foo(%ST* %s) {
3603 %t1 = getelementptr %ST* %s, i32 1 <i>; yields %ST*:%t1</i>
3604 %t2 = getelementptr %ST* %t1, i32 0, i32 2 <i>; yields %RT*:%t2</i>
3605 %t3 = getelementptr %RT* %t2, i32 0, i32 1 <i>; yields [10 x [20 x i32]]*:%t3</i>
3606 %t4 = getelementptr [10 x [20 x i32]]* %t3, i32 0, i32 5 <i>; yields [20 x i32]*:%t4</i>
3607 %t5 = getelementptr [20 x i32]* %t4, i32 0, i32 13 <i>; yields i32*:%t5</i>
3612 <p>Note that it is undefined to access an array out of bounds: array and
3613 pointer indexes must always be within the defined bounds of the array type.
3614 The one exception for this rule is zero length arrays. These arrays are
3615 defined to be accessible as variable length arrays, which requires access
3616 beyond the zero'th element.</p>
3618 <p>The getelementptr instruction is often confusing. For some more insight
3619 into how it works, see <a href="GetElementPtr.html">the getelementptr
3625 <i>; yields [12 x i8]*:aptr</i>
3626 %aptr = getelementptr {i32, [12 x i8]}* %saptr, i64 0, i32 1
3627 <i>; yields i8*:vptr</i>
3628 %vptr = getelementptr {i32, <2 x i8>}* %svptr, i64 0, i32 1, i32 1
3629 <i>; yields i8*:eptr</i>
3630 %eptr = getelementptr [12 x i8]* %aptr, i64 0, i32 1
3634 <!-- ======================================================================= -->
3635 <div class="doc_subsection"> <a name="convertops">Conversion Operations</a>
3637 <div class="doc_text">
3638 <p>The instructions in this category are the conversion instructions (casting)
3639 which all take a single operand and a type. They perform various bit conversions
3643 <!-- _______________________________________________________________________ -->
3644 <div class="doc_subsubsection">
3645 <a name="i_trunc">'<tt>trunc .. to</tt>' Instruction</a>
3647 <div class="doc_text">
3651 <result> = trunc <ty> <value> to <ty2> <i>; yields ty2</i>
3656 The '<tt>trunc</tt>' instruction truncates its operand to the type <tt>ty2</tt>.
3661 The '<tt>trunc</tt>' instruction takes a <tt>value</tt> to trunc, which must
3662 be an <a href="#t_integer">integer</a> type, and a type that specifies the size
3663 and type of the result, which must be an <a href="#t_integer">integer</a>
3664 type. The bit size of <tt>value</tt> must be larger than the bit size of
3665 <tt>ty2</tt>. Equal sized types are not allowed.</p>
3669 The '<tt>trunc</tt>' instruction truncates the high order bits in <tt>value</tt>
3670 and converts the remaining bits to <tt>ty2</tt>. Since the source size must be
3671 larger than the destination size, <tt>trunc</tt> cannot be a <i>no-op cast</i>.
3672 It will always truncate bits.</p>
3676 %X = trunc i32 257 to i8 <i>; yields i8:1</i>
3677 %Y = trunc i32 123 to i1 <i>; yields i1:true</i>
3678 %Y = trunc i32 122 to i1 <i>; yields i1:false</i>
3682 <!-- _______________________________________________________________________ -->
3683 <div class="doc_subsubsection">
3684 <a name="i_zext">'<tt>zext .. to</tt>' Instruction</a>
3686 <div class="doc_text">
3690 <result> = zext <ty> <value> to <ty2> <i>; yields ty2</i>
3694 <p>The '<tt>zext</tt>' instruction zero extends its operand to type
3699 <p>The '<tt>zext</tt>' instruction takes a value to cast, which must be of
3700 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
3701 also be of <a href="#t_integer">integer</a> type. The bit size of the
3702 <tt>value</tt> must be smaller than the bit size of the destination type,
3706 <p>The <tt>zext</tt> fills the high order bits of the <tt>value</tt> with zero
3707 bits until it reaches the size of the destination type, <tt>ty2</tt>.</p>
3709 <p>When zero extending from i1, the result will always be either 0 or 1.</p>
3713 %X = zext i32 257 to i64 <i>; yields i64:257</i>
3714 %Y = zext i1 true to i32 <i>; yields i32:1</i>
3718 <!-- _______________________________________________________________________ -->
3719 <div class="doc_subsubsection">
3720 <a name="i_sext">'<tt>sext .. to</tt>' Instruction</a>
3722 <div class="doc_text">
3726 <result> = sext <ty> <value> to <ty2> <i>; yields ty2</i>
3730 <p>The '<tt>sext</tt>' sign extends <tt>value</tt> to the type <tt>ty2</tt>.</p>
3734 The '<tt>sext</tt>' instruction takes a value to cast, which must be of
3735 <a href="#t_integer">integer</a> type, and a type to cast it to, which must
3736 also be of <a href="#t_integer">integer</a> type. The bit size of the
3737 <tt>value</tt> must be smaller than the bit size of the destination type,
3742 The '<tt>sext</tt>' instruction performs a sign extension by copying the sign
3743 bit (highest order bit) of the <tt>value</tt> until it reaches the bit size of
3744 the type <tt>ty2</tt>.</p>
3746 <p>When sign extending from i1, the extension always results in -1 or 0.</p>
3750 %X = sext i8 -1 to i16 <i>; yields i16 :65535</i>
3751 %Y = sext i1 true to i32 <i>; yields i32:-1</i>
3755 <!-- _______________________________________________________________________ -->
3756 <div class="doc_subsubsection">
3757 <a name="i_fptrunc">'<tt>fptrunc .. to</tt>' Instruction</a>
3760 <div class="doc_text">
3765 <result> = fptrunc <ty> <value> to <ty2> <i>; yields ty2</i>
3769 <p>The '<tt>fptrunc</tt>' instruction truncates <tt>value</tt> to type
3774 <p>The '<tt>fptrunc</tt>' instruction takes a <a href="#t_floating">floating
3775 point</a> value to cast and a <a href="#t_floating">floating point</a> type to
3776 cast it to. The size of <tt>value</tt> must be larger than the size of
3777 <tt>ty2</tt>. This implies that <tt>fptrunc</tt> cannot be used to make a
3778 <i>no-op cast</i>.</p>
3781 <p> The '<tt>fptrunc</tt>' instruction truncates a <tt>value</tt> from a larger
3782 <a href="#t_floating">floating point</a> type to a smaller
3783 <a href="#t_floating">floating point</a> type. If the value cannot fit within
3784 the destination type, <tt>ty2</tt>, then the results are undefined.</p>
3788 %X = fptrunc double 123.0 to float <i>; yields float:123.0</i>
3789 %Y = fptrunc double 1.0E+300 to float <i>; yields undefined</i>
3793 <!-- _______________________________________________________________________ -->
3794 <div class="doc_subsubsection">
3795 <a name="i_fpext">'<tt>fpext .. to</tt>' Instruction</a>
3797 <div class="doc_text">
3801 <result> = fpext <ty> <value> to <ty2> <i>; yields ty2</i>
3805 <p>The '<tt>fpext</tt>' extends a floating point <tt>value</tt> to a larger
3806 floating point value.</p>
3809 <p>The '<tt>fpext</tt>' instruction takes a
3810 <a href="#t_floating">floating point</a> <tt>value</tt> to cast,
3811 and a <a href="#t_floating">floating point</a> type to cast it to. The source
3812 type must be smaller than the destination type.</p>
3815 <p>The '<tt>fpext</tt>' instruction extends the <tt>value</tt> from a smaller
3816 <a href="#t_floating">floating point</a> type to a larger
3817 <a href="#t_floating">floating point</a> type. The <tt>fpext</tt> cannot be
3818 used to make a <i>no-op cast</i> because it always changes bits. Use
3819 <tt>bitcast</tt> to make a <i>no-op cast</i> for a floating point cast.</p>
3823 %X = fpext float 3.1415 to double <i>; yields double:3.1415</i>
3824 %Y = fpext float 1.0 to float <i>; yields float:1.0 (no-op)</i>
3828 <!-- _______________________________________________________________________ -->
3829 <div class="doc_subsubsection">
3830 <a name="i_fptoui">'<tt>fptoui .. to</tt>' Instruction</a>
3832 <div class="doc_text">
3836 <result> = fptoui <ty> <value> to <ty2> <i>; yields ty2</i>
3840 <p>The '<tt>fptoui</tt>' converts a floating point <tt>value</tt> to its
3841 unsigned integer equivalent of type <tt>ty2</tt>.
3845 <p>The '<tt>fptoui</tt>' instruction takes a value to cast, which must be a
3846 scalar or vector <a href="#t_floating">floating point</a> value, and a type
3847 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3848 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3849 vector integer type with the same number of elements as <tt>ty</tt></p>
3852 <p> The '<tt>fptoui</tt>' instruction converts its
3853 <a href="#t_floating">floating point</a> operand into the nearest (rounding
3854 towards zero) unsigned integer value. If the value cannot fit in <tt>ty2</tt>,
3855 the results are undefined.</p>
3859 %X = fptoui double 123.0 to i32 <i>; yields i32:123</i>
3860 %Y = fptoui float 1.0E+300 to i1 <i>; yields undefined:1</i>
3861 %X = fptoui float 1.04E+17 to i8 <i>; yields undefined:1</i>
3865 <!-- _______________________________________________________________________ -->
3866 <div class="doc_subsubsection">
3867 <a name="i_fptosi">'<tt>fptosi .. to</tt>' Instruction</a>
3869 <div class="doc_text">
3873 <result> = fptosi <ty> <value> to <ty2> <i>; yields ty2</i>
3877 <p>The '<tt>fptosi</tt>' instruction converts
3878 <a href="#t_floating">floating point</a> <tt>value</tt> to type <tt>ty2</tt>.
3882 <p> The '<tt>fptosi</tt>' instruction takes a value to cast, which must be a
3883 scalar or vector <a href="#t_floating">floating point</a> value, and a type
3884 to cast it to <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a>
3885 type. If <tt>ty</tt> is a vector floating point type, <tt>ty2</tt> must be a
3886 vector integer type with the same number of elements as <tt>ty</tt></p>
3889 <p>The '<tt>fptosi</tt>' instruction converts its
3890 <a href="#t_floating">floating point</a> operand into the nearest (rounding
3891 towards zero) signed integer value. If the value cannot fit in <tt>ty2</tt>,
3892 the results are undefined.</p>
3896 %X = fptosi double -123.0 to i32 <i>; yields i32:-123</i>
3897 %Y = fptosi float 1.0E-247 to i1 <i>; yields undefined:1</i>
3898 %X = fptosi float 1.04E+17 to i8 <i>; yields undefined:1</i>
3902 <!-- _______________________________________________________________________ -->
3903 <div class="doc_subsubsection">
3904 <a name="i_uitofp">'<tt>uitofp .. to</tt>' Instruction</a>
3906 <div class="doc_text">
3910 <result> = uitofp <ty> <value> to <ty2> <i>; yields ty2</i>
3914 <p>The '<tt>uitofp</tt>' instruction regards <tt>value</tt> as an unsigned
3915 integer and converts that value to the <tt>ty2</tt> type.</p>
3918 <p>The '<tt>uitofp</tt>' instruction takes a value to cast, which must be a
3919 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3920 to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3921 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3922 floating point type with the same number of elements as <tt>ty</tt></p>
3925 <p>The '<tt>uitofp</tt>' instruction interprets its operand as an unsigned
3926 integer quantity and converts it to the corresponding floating point value. If
3927 the value cannot fit in the floating point value, the results are undefined.</p>
3931 %X = uitofp i32 257 to float <i>; yields float:257.0</i>
3932 %Y = uitofp i8 -1 to double <i>; yields double:255.0</i>
3936 <!-- _______________________________________________________________________ -->
3937 <div class="doc_subsubsection">
3938 <a name="i_sitofp">'<tt>sitofp .. to</tt>' Instruction</a>
3940 <div class="doc_text">
3944 <result> = sitofp <ty> <value> to <ty2> <i>; yields ty2</i>
3948 <p>The '<tt>sitofp</tt>' instruction regards <tt>value</tt> as a signed
3949 integer and converts that value to the <tt>ty2</tt> type.</p>
3952 <p>The '<tt>sitofp</tt>' instruction takes a value to cast, which must be a
3953 scalar or vector <a href="#t_integer">integer</a> value, and a type to cast it
3954 to <tt>ty2</tt>, which must be an <a href="#t_floating">floating point</a>
3955 type. If <tt>ty</tt> is a vector integer type, <tt>ty2</tt> must be a vector
3956 floating point type with the same number of elements as <tt>ty</tt></p>
3959 <p>The '<tt>sitofp</tt>' instruction interprets its operand as a signed
3960 integer quantity and converts it to the corresponding floating point value. If
3961 the value cannot fit in the floating point value, the results are undefined.</p>
3965 %X = sitofp i32 257 to float <i>; yields float:257.0</i>
3966 %Y = sitofp i8 -1 to double <i>; yields double:-1.0</i>
3970 <!-- _______________________________________________________________________ -->
3971 <div class="doc_subsubsection">
3972 <a name="i_ptrtoint">'<tt>ptrtoint .. to</tt>' Instruction</a>
3974 <div class="doc_text">
3978 <result> = ptrtoint <ty> <value> to <ty2> <i>; yields ty2</i>
3982 <p>The '<tt>ptrtoint</tt>' instruction converts the pointer <tt>value</tt> to
3983 the integer type <tt>ty2</tt>.</p>
3986 <p>The '<tt>ptrtoint</tt>' instruction takes a <tt>value</tt> to cast, which
3987 must be a <a href="#t_pointer">pointer</a> value, and a type to cast it to
3988 <tt>ty2</tt>, which must be an <a href="#t_integer">integer</a> type.</p>
3991 <p>The '<tt>ptrtoint</tt>' instruction converts <tt>value</tt> to integer type
3992 <tt>ty2</tt> by interpreting the pointer value as an integer and either
3993 truncating or zero extending that value to the size of the integer type. If
3994 <tt>value</tt> is smaller than <tt>ty2</tt> then a zero extension is done. If
3995 <tt>value</tt> is larger than <tt>ty2</tt> then a truncation is done. If they
3996 are the same size, then nothing is done (<i>no-op cast</i>) other than a type
4001 %X = ptrtoint i32* %X to i8 <i>; yields truncation on 32-bit architecture</i>
4002 %Y = ptrtoint i32* %x to i64 <i>; yields zero extension on 32-bit architecture</i>
4006 <!-- _______________________________________________________________________ -->
4007 <div class="doc_subsubsection">
4008 <a name="i_inttoptr">'<tt>inttoptr .. to</tt>' Instruction</a>
4010 <div class="doc_text">
4014 <result> = inttoptr <ty> <value> to <ty2> <i>; yields ty2</i>
4018 <p>The '<tt>inttoptr</tt>' instruction converts an integer <tt>value</tt> to
4019 a pointer type, <tt>ty2</tt>.</p>
4022 <p>The '<tt>inttoptr</tt>' instruction takes an <a href="#t_integer">integer</a>
4023 value to cast, and a type to cast it to, which must be a
4024 <a href="#t_pointer">pointer</a> type.</p>
4027 <p>The '<tt>inttoptr</tt>' instruction converts <tt>value</tt> to type
4028 <tt>ty2</tt> by applying either a zero extension or a truncation depending on
4029 the size of the integer <tt>value</tt>. If <tt>value</tt> is larger than the
4030 size of a pointer then a truncation is done. If <tt>value</tt> is smaller than
4031 the size of a pointer then a zero extension is done. If they are the same size,
4032 nothing is done (<i>no-op cast</i>).</p>
4036 %X = inttoptr i32 255 to i32* <i>; yields zero extension on 64-bit architecture</i>
4037 %X = inttoptr i32 255 to i32* <i>; yields no-op on 32-bit architecture</i>
4038 %Y = inttoptr i64 0 to i32* <i>; yields truncation on 32-bit architecture</i>
4042 <!-- _______________________________________________________________________ -->
4043 <div class="doc_subsubsection">
4044 <a name="i_bitcast">'<tt>bitcast .. to</tt>' Instruction</a>
4046 <div class="doc_text">
4050 <result> = bitcast <ty> <value> to <ty2> <i>; yields ty2</i>
4055 <p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
4056 <tt>ty2</tt> without changing any bits.</p>
4060 <p>The '<tt>bitcast</tt>' instruction takes a value to cast, which must be
4061 a non-aggregate first class value, and a type to cast it to, which must also be
4062 a non-aggregate <a href="#t_firstclass">first class</a> type. The bit sizes of
4064 and the destination type, <tt>ty2</tt>, must be identical. If the source
4065 type is a pointer, the destination type must also be a pointer. This
4066 instruction supports bitwise conversion of vectors to integers and to vectors
4067 of other types (as long as they have the same size).</p>
4070 <p>The '<tt>bitcast</tt>' instruction converts <tt>value</tt> to type
4071 <tt>ty2</tt>. It is always a <i>no-op cast</i> because no bits change with
4072 this conversion. The conversion is done as if the <tt>value</tt> had been
4073 stored to memory and read back as type <tt>ty2</tt>. Pointer types may only be
4074 converted to other pointer types with this instruction. To convert pointers to
4075 other types, use the <a href="#i_inttoptr">inttoptr</a> or
4076 <a href="#i_ptrtoint">ptrtoint</a> instructions first.</p>
4080 %X = bitcast i8 255 to i8 <i>; yields i8 :-1</i>
4081 %Y = bitcast i32* %x to sint* <i>; yields sint*:%x</i>
4082 %Z = bitcast <2 x int> %V to i64; <i>; yields i64: %V</i>
4086 <!-- ======================================================================= -->
4087 <div class="doc_subsection"> <a name="otherops">Other Operations</a> </div>
4088 <div class="doc_text">
4089 <p>The instructions in this category are the "miscellaneous"
4090 instructions, which defy better classification.</p>
4093 <!-- _______________________________________________________________________ -->
4094 <div class="doc_subsubsection"><a name="i_icmp">'<tt>icmp</tt>' Instruction</a>
4096 <div class="doc_text">
4098 <pre> <result> = icmp <cond> <ty> <op1>, <op2> <i>; yields {i1} or {<N x i1>}:result</i>
4101 <p>The '<tt>icmp</tt>' instruction returns a boolean value or
4102 a vector of boolean values based on comparison
4103 of its two integer, integer vector, or pointer operands.</p>
4105 <p>The '<tt>icmp</tt>' instruction takes three operands. The first operand is
4106 the condition code indicating the kind of comparison to perform. It is not
4107 a value, just a keyword. The possible condition code are:
4110 <li><tt>eq</tt>: equal</li>
4111 <li><tt>ne</tt>: not equal </li>
4112 <li><tt>ugt</tt>: unsigned greater than</li>
4113 <li><tt>uge</tt>: unsigned greater or equal</li>
4114 <li><tt>ult</tt>: unsigned less than</li>
4115 <li><tt>ule</tt>: unsigned less or equal</li>
4116 <li><tt>sgt</tt>: signed greater than</li>
4117 <li><tt>sge</tt>: signed greater or equal</li>
4118 <li><tt>slt</tt>: signed less than</li>
4119 <li><tt>sle</tt>: signed less or equal</li>
4121 <p>The remaining two arguments must be <a href="#t_integer">integer</a> or
4122 <a href="#t_pointer">pointer</a>
4123 or integer <a href="#t_vector">vector</a> typed.
4124 They must also be identical types.</p>
4126 <p>The '<tt>icmp</tt>' compares <tt>op1</tt> and <tt>op2</tt> according to
4127 the condition code given as <tt>cond</tt>. The comparison performed always
4128 yields either an <a href="#t_primitive"><tt>i1</tt></a> or vector of <tt>i1</tt> result, as follows:
4131 <li><tt>eq</tt>: yields <tt>true</tt> if the operands are equal,
4132 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.
4134 <li><tt>ne</tt>: yields <tt>true</tt> if the operands are unequal,
4135 <tt>false</tt> otherwise. No sign interpretation is necessary or performed.</li>
4136 <li><tt>ugt</tt>: interprets the operands as unsigned values and yields
4137 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4138 <li><tt>uge</tt>: interprets the operands as unsigned values and yields
4139 <tt>true</tt> if <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
4140 <li><tt>ult</tt>: interprets the operands as unsigned values and yields
4141 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
4142 <li><tt>ule</tt>: interprets the operands as unsigned values and yields
4143 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4144 <li><tt>sgt</tt>: interprets the operands as signed values and yields
4145 <tt>true</tt> if <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4146 <li><tt>sge</tt>: interprets the operands as signed values and yields
4147 <tt>true</tt> if <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
4148 <li><tt>slt</tt>: interprets the operands as signed values and yields
4149 <tt>true</tt> if <tt>op1</tt> is less than <tt>op2</tt>.</li>
4150 <li><tt>sle</tt>: interprets the operands as signed values and yields
4151 <tt>true</tt> if <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4153 <p>If the operands are <a href="#t_pointer">pointer</a> typed, the pointer
4154 values are compared as if they were integers.</p>
4155 <p>If the operands are integer vectors, then they are compared
4156 element by element. The result is an <tt>i1</tt> vector with
4157 the same number of elements as the values being compared.
4158 Otherwise, the result is an <tt>i1</tt>.
4162 <pre> <result> = icmp eq i32 4, 5 <i>; yields: result=false</i>
4163 <result> = icmp ne float* %X, %X <i>; yields: result=false</i>
4164 <result> = icmp ult i16 4, 5 <i>; yields: result=true</i>
4165 <result> = icmp sgt i16 4, 5 <i>; yields: result=false</i>
4166 <result> = icmp ule i16 -4, 5 <i>; yields: result=false</i>
4167 <result> = icmp sge i16 4, 5 <i>; yields: result=false</i>
4170 <p>Note that the code generator does not yet support vector types with
4171 the <tt>icmp</tt> instruction.</p>
4175 <!-- _______________________________________________________________________ -->
4176 <div class="doc_subsubsection"><a name="i_fcmp">'<tt>fcmp</tt>' Instruction</a>
4178 <div class="doc_text">
4180 <pre> <result> = fcmp <cond> <ty> <op1>, <op2> <i>; yields {i1} or {<N x i1>}:result</i>
4183 <p>The '<tt>fcmp</tt>' instruction returns a boolean value
4184 or vector of boolean values based on comparison
4185 of its operands.</p>
4187 If the operands are floating point scalars, then the result
4188 type is a boolean (<a href="#t_primitive"><tt>i1</tt></a>).
4190 <p>If the operands are floating point vectors, then the result type
4191 is a vector of boolean with the same number of elements as the
4192 operands being compared.</p>
4194 <p>The '<tt>fcmp</tt>' instruction takes three operands. The first operand is
4195 the condition code indicating the kind of comparison to perform. It is not
4196 a value, just a keyword. The possible condition code are:</p>
4198 <li><tt>false</tt>: no comparison, always returns false</li>
4199 <li><tt>oeq</tt>: ordered and equal</li>
4200 <li><tt>ogt</tt>: ordered and greater than </li>
4201 <li><tt>oge</tt>: ordered and greater than or equal</li>
4202 <li><tt>olt</tt>: ordered and less than </li>
4203 <li><tt>ole</tt>: ordered and less than or equal</li>
4204 <li><tt>one</tt>: ordered and not equal</li>
4205 <li><tt>ord</tt>: ordered (no nans)</li>
4206 <li><tt>ueq</tt>: unordered or equal</li>
4207 <li><tt>ugt</tt>: unordered or greater than </li>
4208 <li><tt>uge</tt>: unordered or greater than or equal</li>
4209 <li><tt>ult</tt>: unordered or less than </li>
4210 <li><tt>ule</tt>: unordered or less than or equal</li>
4211 <li><tt>une</tt>: unordered or not equal</li>
4212 <li><tt>uno</tt>: unordered (either nans)</li>
4213 <li><tt>true</tt>: no comparison, always returns true</li>
4215 <p><i>Ordered</i> means that neither operand is a QNAN while
4216 <i>unordered</i> means that either operand may be a QNAN.</p>
4217 <p>Each of <tt>val1</tt> and <tt>val2</tt> arguments must be
4218 either a <a href="#t_floating">floating point</a> type
4219 or a <a href="#t_vector">vector</a> of floating point type.
4220 They must have identical types.</p>
4222 <p>The '<tt>fcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
4223 according to the condition code given as <tt>cond</tt>.
4224 If the operands are vectors, then the vectors are compared
4226 Each comparison performed
4227 always yields an <a href="#t_primitive">i1</a> result, as follows:</p>
4229 <li><tt>false</tt>: always yields <tt>false</tt>, regardless of operands.</li>
4230 <li><tt>oeq</tt>: yields <tt>true</tt> if both operands are not a QNAN and
4231 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
4232 <li><tt>ogt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
4233 <tt>op1</tt> is greather than <tt>op2</tt>.</li>
4234 <li><tt>oge</tt>: yields <tt>true</tt> if both operands are not a QNAN and
4235 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
4236 <li><tt>olt</tt>: yields <tt>true</tt> if both operands are not a QNAN and
4237 <tt>op1</tt> is less than <tt>op2</tt>.</li>
4238 <li><tt>ole</tt>: yields <tt>true</tt> if both operands are not a QNAN and
4239 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4240 <li><tt>one</tt>: yields <tt>true</tt> if both operands are not a QNAN and
4241 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
4242 <li><tt>ord</tt>: yields <tt>true</tt> if both operands are not a QNAN.</li>
4243 <li><tt>ueq</tt>: yields <tt>true</tt> if either operand is a QNAN or
4244 <tt>op1</tt> is equal to <tt>op2</tt>.</li>
4245 <li><tt>ugt</tt>: yields <tt>true</tt> if either operand is a QNAN or
4246 <tt>op1</tt> is greater than <tt>op2</tt>.</li>
4247 <li><tt>uge</tt>: yields <tt>true</tt> if either operand is a QNAN or
4248 <tt>op1</tt> is greater than or equal to <tt>op2</tt>.</li>
4249 <li><tt>ult</tt>: yields <tt>true</tt> if either operand is a QNAN or
4250 <tt>op1</tt> is less than <tt>op2</tt>.</li>
4251 <li><tt>ule</tt>: yields <tt>true</tt> if either operand is a QNAN or
4252 <tt>op1</tt> is less than or equal to <tt>op2</tt>.</li>
4253 <li><tt>une</tt>: yields <tt>true</tt> if either operand is a QNAN or
4254 <tt>op1</tt> is not equal to <tt>op2</tt>.</li>
4255 <li><tt>uno</tt>: yields <tt>true</tt> if either operand is a QNAN.</li>
4256 <li><tt>true</tt>: always yields <tt>true</tt>, regardless of operands.</li>
4260 <pre> <result> = fcmp oeq float 4.0, 5.0 <i>; yields: result=false</i>
4261 <result> = fcmp one float 4.0, 5.0 <i>; yields: result=true</i>
4262 <result> = fcmp olt float 4.0, 5.0 <i>; yields: result=true</i>
4263 <result> = fcmp ueq double 1.0, 2.0 <i>; yields: result=false</i>
4266 <p>Note that the code generator does not yet support vector types with
4267 the <tt>fcmp</tt> instruction.</p>
4271 <!-- _______________________________________________________________________ -->
4272 <div class="doc_subsubsection">
4273 <a name="i_vicmp">'<tt>vicmp</tt>' Instruction</a>
4275 <div class="doc_text">
4277 <pre> <result> = vicmp <cond> <ty> <op1>, <op2> <i>; yields {ty}:result</i>
4280 <p>The '<tt>vicmp</tt>' instruction returns an integer vector value based on
4281 element-wise comparison of its two integer vector operands.</p>
4283 <p>The '<tt>vicmp</tt>' instruction takes three operands. The first operand is
4284 the condition code indicating the kind of comparison to perform. It is not
4285 a value, just a keyword. The possible condition code are:</p>
4287 <li><tt>eq</tt>: equal</li>
4288 <li><tt>ne</tt>: not equal </li>
4289 <li><tt>ugt</tt>: unsigned greater than</li>
4290 <li><tt>uge</tt>: unsigned greater or equal</li>
4291 <li><tt>ult</tt>: unsigned less than</li>
4292 <li><tt>ule</tt>: unsigned less or equal</li>
4293 <li><tt>sgt</tt>: signed greater than</li>
4294 <li><tt>sge</tt>: signed greater or equal</li>
4295 <li><tt>slt</tt>: signed less than</li>
4296 <li><tt>sle</tt>: signed less or equal</li>
4298 <p>The remaining two arguments must be <a href="#t_vector">vector</a> or
4299 <a href="#t_integer">integer</a> typed. They must also be identical types.</p>
4301 <p>The '<tt>vicmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
4302 according to the condition code given as <tt>cond</tt>. The comparison yields a
4303 <a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, of
4304 identical type as the values being compared. The most significant bit in each
4305 element is 1 if the element-wise comparison evaluates to true, and is 0
4306 otherwise. All other bits of the result are undefined. The condition codes
4307 are evaluated identically to the <a href="#i_icmp">'<tt>icmp</tt>'
4308 instruction</a>.</p>
4312 <result> = vicmp eq <2 x i32> < i32 4, i32 0>, < i32 5, i32 0> <i>; yields: result=<2 x i32> < i32 0, i32 -1 ></i>
4313 <result> = vicmp ult <2 x i8 > < i8 1, i8 2>, < i8 2, i8 2 > <i>; yields: result=<2 x i8> < i8 -1, i8 0 ></i>
4317 <!-- _______________________________________________________________________ -->
4318 <div class="doc_subsubsection">
4319 <a name="i_vfcmp">'<tt>vfcmp</tt>' Instruction</a>
4321 <div class="doc_text">
4323 <pre> <result> = vfcmp <cond> <ty> <op1>, <op2></pre>
4325 <p>The '<tt>vfcmp</tt>' instruction returns an integer vector value based on
4326 element-wise comparison of its two floating point vector operands. The output
4327 elements have the same width as the input elements.</p>
4329 <p>The '<tt>vfcmp</tt>' instruction takes three operands. The first operand is
4330 the condition code indicating the kind of comparison to perform. It is not
4331 a value, just a keyword. The possible condition code are:</p>
4333 <li><tt>false</tt>: no comparison, always returns false</li>
4334 <li><tt>oeq</tt>: ordered and equal</li>
4335 <li><tt>ogt</tt>: ordered and greater than </li>
4336 <li><tt>oge</tt>: ordered and greater than or equal</li>
4337 <li><tt>olt</tt>: ordered and less than </li>
4338 <li><tt>ole</tt>: ordered and less than or equal</li>
4339 <li><tt>one</tt>: ordered and not equal</li>
4340 <li><tt>ord</tt>: ordered (no nans)</li>
4341 <li><tt>ueq</tt>: unordered or equal</li>
4342 <li><tt>ugt</tt>: unordered or greater than </li>
4343 <li><tt>uge</tt>: unordered or greater than or equal</li>
4344 <li><tt>ult</tt>: unordered or less than </li>
4345 <li><tt>ule</tt>: unordered or less than or equal</li>
4346 <li><tt>une</tt>: unordered or not equal</li>
4347 <li><tt>uno</tt>: unordered (either nans)</li>
4348 <li><tt>true</tt>: no comparison, always returns true</li>
4350 <p>The remaining two arguments must be <a href="#t_vector">vector</a> of
4351 <a href="#t_floating">floating point</a> typed. They must also be identical
4354 <p>The '<tt>vfcmp</tt>' instruction compares <tt>op1</tt> and <tt>op2</tt>
4355 according to the condition code given as <tt>cond</tt>. The comparison yields a
4356 <a href="#t_vector">vector</a> of <a href="#t_integer">integer</a> result, with
4357 an identical number of elements as the values being compared, and each element
4358 having identical with to the width of the floating point elements. The most
4359 significant bit in each element is 1 if the element-wise comparison evaluates to
4360 true, and is 0 otherwise. All other bits of the result are undefined. The
4361 condition codes are evaluated identically to the
4362 <a href="#i_fcmp">'<tt>fcmp</tt>' instruction</a>.</p>
4366 <i>; yields: result=<2 x i32> < i32 0, i32 -1 ></i>
4367 <result> = vfcmp oeq <2 x float> < float 4, float 0 >, < float 5, float 0 >
4369 <i>; yields: result=<2 x i64> < i64 -1, i64 0 ></i>
4370 <result> = vfcmp ult <2 x double> < double 1, double 2 >, < double 2, double 2>
4374 <!-- _______________________________________________________________________ -->
4375 <div class="doc_subsubsection">
4376 <a name="i_phi">'<tt>phi</tt>' Instruction</a>
4379 <div class="doc_text">
4383 <pre> <result> = phi <ty> [ <val0>, <label0>], ...<br></pre>
4385 <p>The '<tt>phi</tt>' instruction is used to implement the φ node in
4386 the SSA graph representing the function.</p>
4389 <p>The type of the incoming values is specified with the first type
4390 field. After this, the '<tt>phi</tt>' instruction takes a list of pairs
4391 as arguments, with one pair for each predecessor basic block of the
4392 current block. Only values of <a href="#t_firstclass">first class</a>
4393 type may be used as the value arguments to the PHI node. Only labels
4394 may be used as the label arguments.</p>
4396 <p>There must be no non-phi instructions between the start of a basic
4397 block and the PHI instructions: i.e. PHI instructions must be first in
4402 <p>At runtime, the '<tt>phi</tt>' instruction logically takes on the value
4403 specified by the pair corresponding to the predecessor basic block that executed
4404 just prior to the current block.</p>
4408 Loop: ; Infinite loop that counts from 0 on up...
4409 %indvar = phi i32 [ 0, %LoopHeader ], [ %nextindvar, %Loop ]
4410 %nextindvar = add i32 %indvar, 1
4415 <!-- _______________________________________________________________________ -->
4416 <div class="doc_subsubsection">
4417 <a name="i_select">'<tt>select</tt>' Instruction</a>
4420 <div class="doc_text">
4425 <result> = select <i>selty</i> <cond>, <ty> <val1>, <ty> <val2> <i>; yields ty</i>
4427 <i>selty</i> is either i1 or {<N x i1>}
4433 The '<tt>select</tt>' instruction is used to choose one value based on a
4434 condition, without branching.
4441 The '<tt>select</tt>' instruction requires an 'i1' value or
4442 a vector of 'i1' values indicating the
4443 condition, and two values of the same <a href="#t_firstclass">first class</a>
4444 type. If the val1/val2 are vectors and
4445 the condition is a scalar, then entire vectors are selected, not
4446 individual elements.
4452 If the condition is an i1 and it evaluates to 1, the instruction returns the first
4453 value argument; otherwise, it returns the second value argument.
4456 If the condition is a vector of i1, then the value arguments must
4457 be vectors of the same size, and the selection is done element
4464 %X = select i1 true, i8 17, i8 42 <i>; yields i8:17</i>
4467 <p>Note that the code generator does not yet support conditions
4468 with vector type.</p>
4473 <!-- _______________________________________________________________________ -->
4474 <div class="doc_subsubsection">
4475 <a name="i_call">'<tt>call</tt>' Instruction</a>
4478 <div class="doc_text">
4482 <result> = [tail] call [<a href="#callingconv">cconv</a>] [<a href="#paramattrs">ret attrs</a>] <ty> [<fnty>*] <fnptrval>(<function args>) [<a href="#fnattrs">fn attrs</a>]
4487 <p>The '<tt>call</tt>' instruction represents a simple function call.</p>
4491 <p>This instruction requires several arguments:</p>
4495 <p>The optional "tail" marker indicates whether the callee function accesses
4496 any allocas or varargs in the caller. If the "tail" marker is present, the
4497 function call is eligible for tail call optimization. Note that calls may
4498 be marked "tail" even if they do not occur before a <a
4499 href="#i_ret"><tt>ret</tt></a> instruction.</p>
4502 <p>The optional "cconv" marker indicates which <a href="#callingconv">calling
4503 convention</a> the call should use. If none is specified, the call defaults
4504 to using C calling conventions.</p>
4508 <p>The optional <a href="#paramattrs">Parameter Attributes</a> list for
4509 return values. Only '<tt>zeroext</tt>', '<tt>signext</tt>',
4510 and '<tt>inreg</tt>' attributes are valid here.</p>
4514 <p>'<tt>ty</tt>': the type of the call instruction itself which is also
4515 the type of the return value. Functions that return no value are marked
4516 <tt><a href="#t_void">void</a></tt>.</p>
4519 <p>'<tt>fnty</tt>': shall be the signature of the pointer to function
4520 value being invoked. The argument types must match the types implied by
4521 this signature. This type can be omitted if the function is not varargs
4522 and if the function type does not return a pointer to a function.</p>
4525 <p>'<tt>fnptrval</tt>': An LLVM value containing a pointer to a function to
4526 be invoked. In most cases, this is a direct function invocation, but
4527 indirect <tt>call</tt>s are just as possible, calling an arbitrary pointer
4528 to function value.</p>
4531 <p>'<tt>function args</tt>': argument list whose types match the
4532 function signature argument types. All arguments must be of
4533 <a href="#t_firstclass">first class</a> type. If the function signature
4534 indicates the function accepts a variable number of arguments, the extra
4535 arguments can be specified.</p>
4538 <p>The optional <a href="#fnattrs">function attributes</a> list. Only
4539 '<tt>noreturn</tt>', '<tt>nounwind</tt>', '<tt>readonly</tt>' and
4540 '<tt>readnone</tt>' attributes are valid here.</p>
4546 <p>The '<tt>call</tt>' instruction is used to cause control flow to
4547 transfer to a specified function, with its incoming arguments bound to
4548 the specified values. Upon a '<tt><a href="#i_ret">ret</a></tt>'
4549 instruction in the called function, control flow continues with the
4550 instruction after the function call, and the return value of the
4551 function is bound to the result argument.</p>
4556 %retval = call i32 @test(i32 %argc)
4557 call i32 (i8 *, ...)* @printf(i8 * %msg, i32 12, i8 42) <i>; yields i32</i>
4558 %X = tail call i32 @foo() <i>; yields i32</i>
4559 %Y = tail call <a href="#callingconv">fastcc</a> i32 @foo() <i>; yields i32</i>
4560 call void %foo(i8 97 signext)
4562 %struct.A = type { i32, i8 }
4563 %r = call %struct.A @foo() <i>; yields { 32, i8 }</i>
4564 %gr = extractvalue %struct.A %r, 0 <i>; yields i32</i>
4565 %gr1 = extractvalue %struct.A %r, 1 <i>; yields i8</i>
4566 %Z = call void @foo() noreturn <i>; indicates that %foo never returns normally</i>
4567 %ZZ = call zeroext i32 @bar() <i>; Return value is %zero extended</i>
4572 <!-- _______________________________________________________________________ -->
4573 <div class="doc_subsubsection">
4574 <a name="i_va_arg">'<tt>va_arg</tt>' Instruction</a>
4577 <div class="doc_text">
4582 <resultval> = va_arg <va_list*> <arglist>, <argty>
4587 <p>The '<tt>va_arg</tt>' instruction is used to access arguments passed through
4588 the "variable argument" area of a function call. It is used to implement the
4589 <tt>va_arg</tt> macro in C.</p>
4593 <p>This instruction takes a <tt>va_list*</tt> value and the type of
4594 the argument. It returns a value of the specified argument type and
4595 increments the <tt>va_list</tt> to point to the next argument. The
4596 actual type of <tt>va_list</tt> is target specific.</p>
4600 <p>The '<tt>va_arg</tt>' instruction loads an argument of the specified
4601 type from the specified <tt>va_list</tt> and causes the
4602 <tt>va_list</tt> to point to the next argument. For more information,
4603 see the variable argument handling <a href="#int_varargs">Intrinsic
4606 <p>It is legal for this instruction to be called in a function which does not
4607 take a variable number of arguments, for example, the <tt>vfprintf</tt>
4610 <p><tt>va_arg</tt> is an LLVM instruction instead of an <a
4611 href="#intrinsics">intrinsic function</a> because it takes a type as an
4616 <p>See the <a href="#int_varargs">variable argument processing</a> section.</p>
4618 <p>Note that the code generator does not yet fully support va_arg
4619 on many targets. Also, it does not currently support va_arg with
4620 aggregate types on any target.</p>
4624 <!-- *********************************************************************** -->
4625 <div class="doc_section"> <a name="intrinsics">Intrinsic Functions</a> </div>
4626 <!-- *********************************************************************** -->
4628 <div class="doc_text">
4630 <p>LLVM supports the notion of an "intrinsic function". These functions have
4631 well known names and semantics and are required to follow certain restrictions.
4632 Overall, these intrinsics represent an extension mechanism for the LLVM
4633 language that does not require changing all of the transformations in LLVM when
4634 adding to the language (or the bitcode reader/writer, the parser, etc...).</p>
4636 <p>Intrinsic function names must all start with an "<tt>llvm.</tt>" prefix. This
4637 prefix is reserved in LLVM for intrinsic names; thus, function names may not
4638 begin with this prefix. Intrinsic functions must always be external functions:
4639 you cannot define the body of intrinsic functions. Intrinsic functions may
4640 only be used in call or invoke instructions: it is illegal to take the address
4641 of an intrinsic function. Additionally, because intrinsic functions are part
4642 of the LLVM language, it is required if any are added that they be documented
4645 <p>Some intrinsic functions can be overloaded, i.e., the intrinsic represents
4646 a family of functions that perform the same operation but on different data
4647 types. Because LLVM can represent over 8 million different integer types,
4648 overloading is used commonly to allow an intrinsic function to operate on any
4649 integer type. One or more of the argument types or the result type can be
4650 overloaded to accept any integer type. Argument types may also be defined as
4651 exactly matching a previous argument's type or the result type. This allows an
4652 intrinsic function which accepts multiple arguments, but needs all of them to
4653 be of the same type, to only be overloaded with respect to a single argument or
4656 <p>Overloaded intrinsics will have the names of its overloaded argument types
4657 encoded into its function name, each preceded by a period. Only those types
4658 which are overloaded result in a name suffix. Arguments whose type is matched
4659 against another type do not. For example, the <tt>llvm.ctpop</tt> function can
4660 take an integer of any width and returns an integer of exactly the same integer
4661 width. This leads to a family of functions such as
4662 <tt>i8 @llvm.ctpop.i8(i8 %val)</tt> and <tt>i29 @llvm.ctpop.i29(i29 %val)</tt>.
4663 Only one type, the return type, is overloaded, and only one type suffix is
4664 required. Because the argument's type is matched against the return type, it
4665 does not require its own name suffix.</p>
4667 <p>To learn how to add an intrinsic function, please see the
4668 <a href="ExtendingLLVM.html">Extending LLVM Guide</a>.
4673 <!-- ======================================================================= -->
4674 <div class="doc_subsection">
4675 <a name="int_varargs">Variable Argument Handling Intrinsics</a>
4678 <div class="doc_text">
4680 <p>Variable argument support is defined in LLVM with the <a
4681 href="#i_va_arg"><tt>va_arg</tt></a> instruction and these three
4682 intrinsic functions. These functions are related to the similarly
4683 named macros defined in the <tt><stdarg.h></tt> header file.</p>
4685 <p>All of these functions operate on arguments that use a
4686 target-specific value type "<tt>va_list</tt>". The LLVM assembly
4687 language reference manual does not define what this type is, so all
4688 transformations should be prepared to handle these functions regardless of
4691 <p>This example shows how the <a href="#i_va_arg"><tt>va_arg</tt></a>
4692 instruction and the variable argument handling intrinsic functions are
4695 <div class="doc_code">
4697 define i32 @test(i32 %X, ...) {
4698 ; Initialize variable argument processing
4700 %ap2 = bitcast i8** %ap to i8*
4701 call void @llvm.va_start(i8* %ap2)
4703 ; Read a single integer argument
4704 %tmp = va_arg i8** %ap, i32
4706 ; Demonstrate usage of llvm.va_copy and llvm.va_end
4708 %aq2 = bitcast i8** %aq to i8*
4709 call void @llvm.va_copy(i8* %aq2, i8* %ap2)
4710 call void @llvm.va_end(i8* %aq2)
4712 ; Stop processing of arguments.
4713 call void @llvm.va_end(i8* %ap2)
4717 declare void @llvm.va_start(i8*)
4718 declare void @llvm.va_copy(i8*, i8*)
4719 declare void @llvm.va_end(i8*)
4725 <!-- _______________________________________________________________________ -->
4726 <div class="doc_subsubsection">
4727 <a name="int_va_start">'<tt>llvm.va_start</tt>' Intrinsic</a>
4731 <div class="doc_text">
4733 <pre> declare void %llvm.va_start(i8* <arglist>)<br></pre>
4735 <p>The '<tt>llvm.va_start</tt>' intrinsic initializes
4736 <tt>*<arglist></tt> for subsequent use by <tt><a
4737 href="#i_va_arg">va_arg</a></tt>.</p>
4741 <p>The argument is a pointer to a <tt>va_list</tt> element to initialize.</p>
4745 <p>The '<tt>llvm.va_start</tt>' intrinsic works just like the <tt>va_start</tt>
4746 macro available in C. In a target-dependent way, it initializes the
4747 <tt>va_list</tt> element to which the argument points, so that the next call to
4748 <tt>va_arg</tt> will produce the first variable argument passed to the function.
4749 Unlike the C <tt>va_start</tt> macro, this intrinsic does not need to know the
4750 last argument of the function as the compiler can figure that out.</p>
4754 <!-- _______________________________________________________________________ -->
4755 <div class="doc_subsubsection">
4756 <a name="int_va_end">'<tt>llvm.va_end</tt>' Intrinsic</a>
4759 <div class="doc_text">
4761 <pre> declare void @llvm.va_end(i8* <arglist>)<br></pre>
4764 <p>The '<tt>llvm.va_end</tt>' intrinsic destroys <tt>*<arglist></tt>,
4765 which has been initialized previously with <tt><a href="#int_va_start">llvm.va_start</a></tt>
4766 or <tt><a href="#i_va_copy">llvm.va_copy</a></tt>.</p>
4770 <p>The argument is a pointer to a <tt>va_list</tt> to destroy.</p>
4774 <p>The '<tt>llvm.va_end</tt>' intrinsic works just like the <tt>va_end</tt>
4775 macro available in C. In a target-dependent way, it destroys the
4776 <tt>va_list</tt> element to which the argument points. Calls to <a
4777 href="#int_va_start"><tt>llvm.va_start</tt></a> and <a href="#int_va_copy">
4778 <tt>llvm.va_copy</tt></a> must be matched exactly with calls to
4779 <tt>llvm.va_end</tt>.</p>
4783 <!-- _______________________________________________________________________ -->
4784 <div class="doc_subsubsection">
4785 <a name="int_va_copy">'<tt>llvm.va_copy</tt>' Intrinsic</a>
4788 <div class="doc_text">
4793 declare void @llvm.va_copy(i8* <destarglist>, i8* <srcarglist>)
4798 <p>The '<tt>llvm.va_copy</tt>' intrinsic copies the current argument position
4799 from the source argument list to the destination argument list.</p>
4803 <p>The first argument is a pointer to a <tt>va_list</tt> element to initialize.
4804 The second argument is a pointer to a <tt>va_list</tt> element to copy from.</p>
4809 <p>The '<tt>llvm.va_copy</tt>' intrinsic works just like the <tt>va_copy</tt>
4810 macro available in C. In a target-dependent way, it copies the source
4811 <tt>va_list</tt> element into the destination <tt>va_list</tt> element. This
4812 intrinsic is necessary because the <tt><a href="#int_va_start">
4813 llvm.va_start</a></tt> intrinsic may be arbitrarily complex and require, for
4814 example, memory allocation.</p>
4818 <!-- ======================================================================= -->
4819 <div class="doc_subsection">
4820 <a name="int_gc">Accurate Garbage Collection Intrinsics</a>
4823 <div class="doc_text">
4826 LLVM support for <a href="GarbageCollection.html">Accurate Garbage
4827 Collection</a> (GC) requires the implementation and generation of these
4829 These intrinsics allow identification of <a href="#int_gcroot">GC roots on the
4830 stack</a>, as well as garbage collector implementations that require <a
4831 href="#int_gcread">read</a> and <a href="#int_gcwrite">write</a> barriers.
4832 Front-ends for type-safe garbage collected languages should generate these
4833 intrinsics to make use of the LLVM garbage collectors. For more details, see <a
4834 href="GarbageCollection.html">Accurate Garbage Collection with LLVM</a>.
4837 <p>The garbage collection intrinsics only operate on objects in the generic
4838 address space (address space zero).</p>
4842 <!-- _______________________________________________________________________ -->
4843 <div class="doc_subsubsection">
4844 <a name="int_gcroot">'<tt>llvm.gcroot</tt>' Intrinsic</a>
4847 <div class="doc_text">
4852 declare void @llvm.gcroot(i8** %ptrloc, i8* %metadata)
4857 <p>The '<tt>llvm.gcroot</tt>' intrinsic declares the existence of a GC root to
4858 the code generator, and allows some metadata to be associated with it.</p>
4862 <p>The first argument specifies the address of a stack object that contains the
4863 root pointer. The second pointer (which must be either a constant or a global
4864 value address) contains the meta-data to be associated with the root.</p>
4868 <p>At runtime, a call to this intrinsic stores a null pointer into the "ptrloc"
4869 location. At compile-time, the code generator generates information to allow
4870 the runtime to find the pointer at GC safe points. The '<tt>llvm.gcroot</tt>'
4871 intrinsic may only be used in a function which <a href="#gc">specifies a GC
4877 <!-- _______________________________________________________________________ -->
4878 <div class="doc_subsubsection">
4879 <a name="int_gcread">'<tt>llvm.gcread</tt>' Intrinsic</a>
4882 <div class="doc_text">
4887 declare i8* @llvm.gcread(i8* %ObjPtr, i8** %Ptr)
4892 <p>The '<tt>llvm.gcread</tt>' intrinsic identifies reads of references from heap
4893 locations, allowing garbage collector implementations that require read
4898 <p>The second argument is the address to read from, which should be an address
4899 allocated from the garbage collector. The first object is a pointer to the
4900 start of the referenced object, if needed by the language runtime (otherwise
4905 <p>The '<tt>llvm.gcread</tt>' intrinsic has the same semantics as a load
4906 instruction, but may be replaced with substantially more complex code by the
4907 garbage collector runtime, as needed. The '<tt>llvm.gcread</tt>' intrinsic
4908 may only be used in a function which <a href="#gc">specifies a GC
4914 <!-- _______________________________________________________________________ -->
4915 <div class="doc_subsubsection">
4916 <a name="int_gcwrite">'<tt>llvm.gcwrite</tt>' Intrinsic</a>
4919 <div class="doc_text">
4924 declare void @llvm.gcwrite(i8* %P1, i8* %Obj, i8** %P2)
4929 <p>The '<tt>llvm.gcwrite</tt>' intrinsic identifies writes of references to heap
4930 locations, allowing garbage collector implementations that require write
4931 barriers (such as generational or reference counting collectors).</p>
4935 <p>The first argument is the reference to store, the second is the start of the
4936 object to store it to, and the third is the address of the field of Obj to
4937 store to. If the runtime does not require a pointer to the object, Obj may be
4942 <p>The '<tt>llvm.gcwrite</tt>' intrinsic has the same semantics as a store
4943 instruction, but may be replaced with substantially more complex code by the
4944 garbage collector runtime, as needed. The '<tt>llvm.gcwrite</tt>' intrinsic
4945 may only be used in a function which <a href="#gc">specifies a GC
4952 <!-- ======================================================================= -->
4953 <div class="doc_subsection">
4954 <a name="int_codegen">Code Generator Intrinsics</a>
4957 <div class="doc_text">
4959 These intrinsics are provided by LLVM to expose special features that may only
4960 be implemented with code generator support.
4965 <!-- _______________________________________________________________________ -->
4966 <div class="doc_subsubsection">
4967 <a name="int_returnaddress">'<tt>llvm.returnaddress</tt>' Intrinsic</a>
4970 <div class="doc_text">
4974 declare i8 *@llvm.returnaddress(i32 <level>)
4980 The '<tt>llvm.returnaddress</tt>' intrinsic attempts to compute a
4981 target-specific value indicating the return address of the current function
4982 or one of its callers.
4988 The argument to this intrinsic indicates which function to return the address
4989 for. Zero indicates the calling function, one indicates its caller, etc. The
4990 argument is <b>required</b> to be a constant integer value.
4996 The '<tt>llvm.returnaddress</tt>' intrinsic either returns a pointer indicating
4997 the return address of the specified call frame, or zero if it cannot be
4998 identified. The value returned by this intrinsic is likely to be incorrect or 0
4999 for arguments other than zero, so it should only be used for debugging purposes.
5003 Note that calling this intrinsic does not prevent function inlining or other
5004 aggressive transformations, so the value returned may not be that of the obvious
5005 source-language caller.
5010 <!-- _______________________________________________________________________ -->
5011 <div class="doc_subsubsection">
5012 <a name="int_frameaddress">'<tt>llvm.frameaddress</tt>' Intrinsic</a>
5015 <div class="doc_text">
5019 declare i8 *@llvm.frameaddress(i32 <level>)
5025 The '<tt>llvm.frameaddress</tt>' intrinsic attempts to return the
5026 target-specific frame pointer value for the specified stack frame.
5032 The argument to this intrinsic indicates which function to return the frame
5033 pointer for. Zero indicates the calling function, one indicates its caller,
5034 etc. The argument is <b>required</b> to be a constant integer value.
5040 The '<tt>llvm.frameaddress</tt>' intrinsic either returns a pointer indicating
5041 the frame address of the specified call frame, or zero if it cannot be
5042 identified. The value returned by this intrinsic is likely to be incorrect or 0
5043 for arguments other than zero, so it should only be used for debugging purposes.
5047 Note that calling this intrinsic does not prevent function inlining or other
5048 aggressive transformations, so the value returned may not be that of the obvious
5049 source-language caller.
5053 <!-- _______________________________________________________________________ -->
5054 <div class="doc_subsubsection">
5055 <a name="int_stacksave">'<tt>llvm.stacksave</tt>' Intrinsic</a>
5058 <div class="doc_text">
5062 declare i8 *@llvm.stacksave()
5068 The '<tt>llvm.stacksave</tt>' intrinsic is used to remember the current state of
5069 the function stack, for use with <a href="#int_stackrestore">
5070 <tt>llvm.stackrestore</tt></a>. This is useful for implementing language
5071 features like scoped automatic variable sized arrays in C99.
5077 This intrinsic returns a opaque pointer value that can be passed to <a
5078 href="#int_stackrestore"><tt>llvm.stackrestore</tt></a>. When an
5079 <tt>llvm.stackrestore</tt> intrinsic is executed with a value saved from
5080 <tt>llvm.stacksave</tt>, it effectively restores the state of the stack to the
5081 state it was in when the <tt>llvm.stacksave</tt> intrinsic executed. In
5082 practice, this pops any <a href="#i_alloca">alloca</a> blocks from the stack
5083 that were allocated after the <tt>llvm.stacksave</tt> was executed.
5088 <!-- _______________________________________________________________________ -->
5089 <div class="doc_subsubsection">
5090 <a name="int_stackrestore">'<tt>llvm.stackrestore</tt>' Intrinsic</a>
5093 <div class="doc_text">
5097 declare void @llvm.stackrestore(i8 * %ptr)
5103 The '<tt>llvm.stackrestore</tt>' intrinsic is used to restore the state of
5104 the function stack to the state it was in when the corresponding <a
5105 href="#int_stacksave"><tt>llvm.stacksave</tt></a> intrinsic executed. This is
5106 useful for implementing language features like scoped automatic variable sized
5113 See the description for <a href="#int_stacksave"><tt>llvm.stacksave</tt></a>.
5119 <!-- _______________________________________________________________________ -->
5120 <div class="doc_subsubsection">
5121 <a name="int_prefetch">'<tt>llvm.prefetch</tt>' Intrinsic</a>
5124 <div class="doc_text">
5128 declare void @llvm.prefetch(i8* <address>, i32 <rw>, i32 <locality>)
5135 The '<tt>llvm.prefetch</tt>' intrinsic is a hint to the code generator to insert
5136 a prefetch instruction if supported; otherwise, it is a noop. Prefetches have
5138 effect on the behavior of the program but can change its performance
5145 <tt>address</tt> is the address to be prefetched, <tt>rw</tt> is the specifier
5146 determining if the fetch should be for a read (0) or write (1), and
5147 <tt>locality</tt> is a temporal locality specifier ranging from (0) - no
5148 locality, to (3) - extremely local keep in cache. The <tt>rw</tt> and
5149 <tt>locality</tt> arguments must be constant integers.
5155 This intrinsic does not modify the behavior of the program. In particular,
5156 prefetches cannot trap and do not produce a value. On targets that support this
5157 intrinsic, the prefetch can provide hints to the processor cache for better
5163 <!-- _______________________________________________________________________ -->
5164 <div class="doc_subsubsection">
5165 <a name="int_pcmarker">'<tt>llvm.pcmarker</tt>' Intrinsic</a>
5168 <div class="doc_text">
5172 declare void @llvm.pcmarker(i32 <id>)
5179 The '<tt>llvm.pcmarker</tt>' intrinsic is a method to export a Program Counter
5181 code to simulators and other tools. The method is target specific, but it is
5182 expected that the marker will use exported symbols to transmit the PC of the
5184 The marker makes no guarantees that it will remain with any specific instruction
5185 after optimizations. It is possible that the presence of a marker will inhibit
5186 optimizations. The intended use is to be inserted after optimizations to allow
5187 correlations of simulation runs.
5193 <tt>id</tt> is a numerical id identifying the marker.
5199 This intrinsic does not modify the behavior of the program. Backends that do not
5200 support this intrinisic may ignore it.
5205 <!-- _______________________________________________________________________ -->
5206 <div class="doc_subsubsection">
5207 <a name="int_readcyclecounter">'<tt>llvm.readcyclecounter</tt>' Intrinsic</a>
5210 <div class="doc_text">
5214 declare i64 @llvm.readcyclecounter( )
5221 The '<tt>llvm.readcyclecounter</tt>' intrinsic provides access to the cycle
5222 counter register (or similar low latency, high accuracy clocks) on those targets
5223 that support it. On X86, it should map to RDTSC. On Alpha, it should map to RPCC.
5224 As the backing counters overflow quickly (on the order of 9 seconds on alpha), this
5225 should only be used for small timings.
5231 When directly supported, reading the cycle counter should not modify any memory.
5232 Implementations are allowed to either return a application specific value or a
5233 system wide value. On backends without support, this is lowered to a constant 0.
5238 <!-- ======================================================================= -->
5239 <div class="doc_subsection">
5240 <a name="int_libc">Standard C Library Intrinsics</a>
5243 <div class="doc_text">
5245 LLVM provides intrinsics for a few important standard C library functions.
5246 These intrinsics allow source-language front-ends to pass information about the
5247 alignment of the pointer arguments to the code generator, providing opportunity
5248 for more efficient code generation.
5253 <!-- _______________________________________________________________________ -->
5254 <div class="doc_subsubsection">
5255 <a name="int_memcpy">'<tt>llvm.memcpy</tt>' Intrinsic</a>
5258 <div class="doc_text">
5261 <p>This is an overloaded intrinsic. You can use llvm.memcpy on any integer bit
5262 width. Not all targets support all bit widths however.</p>
5264 declare void @llvm.memcpy.i8(i8 * <dest>, i8 * <src>,
5265 i8 <len>, i32 <align>)
5266 declare void @llvm.memcpy.i16(i8 * <dest>, i8 * <src>,
5267 i16 <len>, i32 <align>)
5268 declare void @llvm.memcpy.i32(i8 * <dest>, i8 * <src>,
5269 i32 <len>, i32 <align>)
5270 declare void @llvm.memcpy.i64(i8 * <dest>, i8 * <src>,
5271 i64 <len>, i32 <align>)
5277 The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
5278 location to the destination location.
5282 Note that, unlike the standard libc function, the <tt>llvm.memcpy.*</tt>
5283 intrinsics do not return a value, and takes an extra alignment argument.
5289 The first argument is a pointer to the destination, the second is a pointer to
5290 the source. The third argument is an integer argument
5291 specifying the number of bytes to copy, and the fourth argument is the alignment
5292 of the source and destination locations.
5296 If the call to this intrinisic has an alignment value that is not 0 or 1, then
5297 the caller guarantees that both the source and destination pointers are aligned
5304 The '<tt>llvm.memcpy.*</tt>' intrinsics copy a block of memory from the source
5305 location to the destination location, which are not allowed to overlap. It
5306 copies "len" bytes of memory over. If the argument is known to be aligned to
5307 some boundary, this can be specified as the fourth argument, otherwise it should
5313 <!-- _______________________________________________________________________ -->
5314 <div class="doc_subsubsection">
5315 <a name="int_memmove">'<tt>llvm.memmove</tt>' Intrinsic</a>
5318 <div class="doc_text">
5321 <p>This is an overloaded intrinsic. You can use llvm.memmove on any integer bit
5322 width. Not all targets support all bit widths however.</p>
5324 declare void @llvm.memmove.i8(i8 * <dest>, i8 * <src>,
5325 i8 <len>, i32 <align>)
5326 declare void @llvm.memmove.i16(i8 * <dest>, i8 * <src>,
5327 i16 <len>, i32 <align>)
5328 declare void @llvm.memmove.i32(i8 * <dest>, i8 * <src>,
5329 i32 <len>, i32 <align>)
5330 declare void @llvm.memmove.i64(i8 * <dest>, i8 * <src>,
5331 i64 <len>, i32 <align>)
5337 The '<tt>llvm.memmove.*</tt>' intrinsics move a block of memory from the source
5338 location to the destination location. It is similar to the
5339 '<tt>llvm.memcpy</tt>' intrinsic but allows the two memory locations to overlap.
5343 Note that, unlike the standard libc function, the <tt>llvm.memmove.*</tt>
5344 intrinsics do not return a value, and takes an extra alignment argument.
5350 The first argument is a pointer to the destination, the second is a pointer to
5351 the source. The third argument is an integer argument
5352 specifying the number of bytes to copy, and the fourth argument is the alignment
5353 of the source and destination locations.
5357 If the call to this intrinisic has an alignment value that is not 0 or 1, then
5358 the caller guarantees that the source and destination pointers are aligned to
5365 The '<tt>llvm.memmove.*</tt>' intrinsics copy a block of memory from the source
5366 location to the destination location, which may overlap. It
5367 copies "len" bytes of memory over. If the argument is known to be aligned to
5368 some boundary, this can be specified as the fourth argument, otherwise it should
5374 <!-- _______________________________________________________________________ -->
5375 <div class="doc_subsubsection">
5376 <a name="int_memset">'<tt>llvm.memset.*</tt>' Intrinsics</a>
5379 <div class="doc_text">
5382 <p>This is an overloaded intrinsic. You can use llvm.memset on any integer bit
5383 width. Not all targets support all bit widths however.</p>
5385 declare void @llvm.memset.i8(i8 * <dest>, i8 <val>,
5386 i8 <len>, i32 <align>)
5387 declare void @llvm.memset.i16(i8 * <dest>, i8 <val>,
5388 i16 <len>, i32 <align>)
5389 declare void @llvm.memset.i32(i8 * <dest>, i8 <val>,
5390 i32 <len>, i32 <align>)
5391 declare void @llvm.memset.i64(i8 * <dest>, i8 <val>,
5392 i64 <len>, i32 <align>)
5398 The '<tt>llvm.memset.*</tt>' intrinsics fill a block of memory with a particular
5403 Note that, unlike the standard libc function, the <tt>llvm.memset</tt> intrinsic
5404 does not return a value, and takes an extra alignment argument.
5410 The first argument is a pointer to the destination to fill, the second is the
5411 byte value to fill it with, the third argument is an integer
5412 argument specifying the number of bytes to fill, and the fourth argument is the
5413 known alignment of destination location.
5417 If the call to this intrinisic has an alignment value that is not 0 or 1, then
5418 the caller guarantees that the destination pointer is aligned to that boundary.
5424 The '<tt>llvm.memset.*</tt>' intrinsics fill "len" bytes of memory starting at
5426 destination location. If the argument is known to be aligned to some boundary,
5427 this can be specified as the fourth argument, otherwise it should be set to 0 or
5433 <!-- _______________________________________________________________________ -->
5434 <div class="doc_subsubsection">
5435 <a name="int_sqrt">'<tt>llvm.sqrt.*</tt>' Intrinsic</a>
5438 <div class="doc_text">
5441 <p>This is an overloaded intrinsic. You can use <tt>llvm.sqrt</tt> on any
5442 floating point or vector of floating point type. Not all targets support all
5445 declare float @llvm.sqrt.f32(float %Val)
5446 declare double @llvm.sqrt.f64(double %Val)
5447 declare x86_fp80 @llvm.sqrt.f80(x86_fp80 %Val)
5448 declare fp128 @llvm.sqrt.f128(fp128 %Val)
5449 declare ppc_fp128 @llvm.sqrt.ppcf128(ppc_fp128 %Val)
5455 The '<tt>llvm.sqrt</tt>' intrinsics return the sqrt of the specified operand,
5456 returning the same value as the libm '<tt>sqrt</tt>' functions would. Unlike
5457 <tt>sqrt</tt> in libm, however, <tt>llvm.sqrt</tt> has undefined behavior for
5458 negative numbers other than -0.0 (which allows for better optimization, because
5459 there is no need to worry about errno being set). <tt>llvm.sqrt(-0.0)</tt> is
5460 defined to return -0.0 like IEEE sqrt.
5466 The argument and return value are floating point numbers of the same type.
5472 This function returns the sqrt of the specified operand if it is a nonnegative
5473 floating point number.
5477 <!-- _______________________________________________________________________ -->
5478 <div class="doc_subsubsection">
5479 <a name="int_powi">'<tt>llvm.powi.*</tt>' Intrinsic</a>
5482 <div class="doc_text">
5485 <p>This is an overloaded intrinsic. You can use <tt>llvm.powi</tt> on any
5486 floating point or vector of floating point type. Not all targets support all
5489 declare float @llvm.powi.f32(float %Val, i32 %power)
5490 declare double @llvm.powi.f64(double %Val, i32 %power)
5491 declare x86_fp80 @llvm.powi.f80(x86_fp80 %Val, i32 %power)
5492 declare fp128 @llvm.powi.f128(fp128 %Val, i32 %power)
5493 declare ppc_fp128 @llvm.powi.ppcf128(ppc_fp128 %Val, i32 %power)
5499 The '<tt>llvm.powi.*</tt>' intrinsics return the first operand raised to the
5500 specified (positive or negative) power. The order of evaluation of
5501 multiplications is not defined. When a vector of floating point type is
5502 used, the second argument remains a scalar integer value.
5508 The second argument is an integer power, and the first is a value to raise to
5515 This function returns the first value raised to the second power with an
5516 unspecified sequence of rounding operations.</p>
5519 <!-- _______________________________________________________________________ -->
5520 <div class="doc_subsubsection">
5521 <a name="int_sin">'<tt>llvm.sin.*</tt>' Intrinsic</a>
5524 <div class="doc_text">
5527 <p>This is an overloaded intrinsic. You can use <tt>llvm.sin</tt> on any
5528 floating point or vector of floating point type. Not all targets support all
5531 declare float @llvm.sin.f32(float %Val)
5532 declare double @llvm.sin.f64(double %Val)
5533 declare x86_fp80 @llvm.sin.f80(x86_fp80 %Val)
5534 declare fp128 @llvm.sin.f128(fp128 %Val)
5535 declare ppc_fp128 @llvm.sin.ppcf128(ppc_fp128 %Val)
5541 The '<tt>llvm.sin.*</tt>' intrinsics return the sine of the operand.
5547 The argument and return value are floating point numbers of the same type.
5553 This function returns the sine of the specified operand, returning the
5554 same values as the libm <tt>sin</tt> functions would, and handles error
5555 conditions in the same way.</p>
5558 <!-- _______________________________________________________________________ -->
5559 <div class="doc_subsubsection">
5560 <a name="int_cos">'<tt>llvm.cos.*</tt>' Intrinsic</a>
5563 <div class="doc_text">
5566 <p>This is an overloaded intrinsic. You can use <tt>llvm.cos</tt> on any
5567 floating point or vector of floating point type. Not all targets support all
5570 declare float @llvm.cos.f32(float %Val)
5571 declare double @llvm.cos.f64(double %Val)
5572 declare x86_fp80 @llvm.cos.f80(x86_fp80 %Val)
5573 declare fp128 @llvm.cos.f128(fp128 %Val)
5574 declare ppc_fp128 @llvm.cos.ppcf128(ppc_fp128 %Val)
5580 The '<tt>llvm.cos.*</tt>' intrinsics return the cosine of the operand.
5586 The argument and return value are floating point numbers of the same type.
5592 This function returns the cosine of the specified operand, returning the
5593 same values as the libm <tt>cos</tt> functions would, and handles error
5594 conditions in the same way.</p>
5597 <!-- _______________________________________________________________________ -->
5598 <div class="doc_subsubsection">
5599 <a name="int_pow">'<tt>llvm.pow.*</tt>' Intrinsic</a>
5602 <div class="doc_text">
5605 <p>This is an overloaded intrinsic. You can use <tt>llvm.pow</tt> on any
5606 floating point or vector of floating point type. Not all targets support all
5609 declare float @llvm.pow.f32(float %Val, float %Power)
5610 declare double @llvm.pow.f64(double %Val, double %Power)
5611 declare x86_fp80 @llvm.pow.f80(x86_fp80 %Val, x86_fp80 %Power)
5612 declare fp128 @llvm.pow.f128(fp128 %Val, fp128 %Power)
5613 declare ppc_fp128 @llvm.pow.ppcf128(ppc_fp128 %Val, ppc_fp128 Power)
5619 The '<tt>llvm.pow.*</tt>' intrinsics return the first operand raised to the
5620 specified (positive or negative) power.
5626 The second argument is a floating point power, and the first is a value to
5627 raise to that power.
5633 This function returns the first value raised to the second power,
5635 same values as the libm <tt>pow</tt> functions would, and handles error
5636 conditions in the same way.</p>
5640 <!-- ======================================================================= -->
5641 <div class="doc_subsection">
5642 <a name="int_manip">Bit Manipulation Intrinsics</a>
5645 <div class="doc_text">
5647 LLVM provides intrinsics for a few important bit manipulation operations.
5648 These allow efficient code generation for some algorithms.
5653 <!-- _______________________________________________________________________ -->
5654 <div class="doc_subsubsection">
5655 <a name="int_bswap">'<tt>llvm.bswap.*</tt>' Intrinsics</a>
5658 <div class="doc_text">
5661 <p>This is an overloaded intrinsic function. You can use bswap on any integer
5662 type that is an even number of bytes (i.e. BitWidth % 16 == 0).</p>
5664 declare i16 @llvm.bswap.i16(i16 <id>)
5665 declare i32 @llvm.bswap.i32(i32 <id>)
5666 declare i64 @llvm.bswap.i64(i64 <id>)
5672 The '<tt>llvm.bswap</tt>' family of intrinsics is used to byte swap integer
5673 values with an even number of bytes (positive multiple of 16 bits). These are
5674 useful for performing operations on data that is not in the target's native
5681 The <tt>llvm.bswap.i16</tt> intrinsic returns an i16 value that has the high
5682 and low byte of the input i16 swapped. Similarly, the <tt>llvm.bswap.i32</tt>
5683 intrinsic returns an i32 value that has the four bytes of the input i32
5684 swapped, so that if the input bytes are numbered 0, 1, 2, 3 then the returned
5685 i32 will have its bytes in 3, 2, 1, 0 order. The <tt>llvm.bswap.i48</tt>,
5686 <tt>llvm.bswap.i64</tt> and other intrinsics extend this concept to
5687 additional even-byte lengths (6 bytes, 8 bytes and more, respectively).
5692 <!-- _______________________________________________________________________ -->
5693 <div class="doc_subsubsection">
5694 <a name="int_ctpop">'<tt>llvm.ctpop.*</tt>' Intrinsic</a>
5697 <div class="doc_text">
5700 <p>This is an overloaded intrinsic. You can use llvm.ctpop on any integer bit
5701 width. Not all targets support all bit widths however.</p>
5703 declare i8 @llvm.ctpop.i8(i8 <src>)
5704 declare i16 @llvm.ctpop.i16(i16 <src>)
5705 declare i32 @llvm.ctpop.i32(i32 <src>)
5706 declare i64 @llvm.ctpop.i64(i64 <src>)
5707 declare i256 @llvm.ctpop.i256(i256 <src>)
5713 The '<tt>llvm.ctpop</tt>' family of intrinsics counts the number of bits set in a
5720 The only argument is the value to be counted. The argument may be of any
5721 integer type. The return type must match the argument type.
5727 The '<tt>llvm.ctpop</tt>' intrinsic counts the 1's in a variable.
5731 <!-- _______________________________________________________________________ -->
5732 <div class="doc_subsubsection">
5733 <a name="int_ctlz">'<tt>llvm.ctlz.*</tt>' Intrinsic</a>
5736 <div class="doc_text">
5739 <p>This is an overloaded intrinsic. You can use <tt>llvm.ctlz</tt> on any
5740 integer bit width. Not all targets support all bit widths however.</p>
5742 declare i8 @llvm.ctlz.i8 (i8 <src>)
5743 declare i16 @llvm.ctlz.i16(i16 <src>)
5744 declare i32 @llvm.ctlz.i32(i32 <src>)
5745 declare i64 @llvm.ctlz.i64(i64 <src>)
5746 declare i256 @llvm.ctlz.i256(i256 <src>)
5752 The '<tt>llvm.ctlz</tt>' family of intrinsic functions counts the number of
5753 leading zeros in a variable.
5759 The only argument is the value to be counted. The argument may be of any
5760 integer type. The return type must match the argument type.
5766 The '<tt>llvm.ctlz</tt>' intrinsic counts the leading (most significant) zeros
5767 in a variable. If the src == 0 then the result is the size in bits of the type
5768 of src. For example, <tt>llvm.ctlz(i32 2) = 30</tt>.
5774 <!-- _______________________________________________________________________ -->
5775 <div class="doc_subsubsection">
5776 <a name="int_cttz">'<tt>llvm.cttz.*</tt>' Intrinsic</a>
5779 <div class="doc_text">
5782 <p>This is an overloaded intrinsic. You can use <tt>llvm.cttz</tt> on any
5783 integer bit width. Not all targets support all bit widths however.</p>
5785 declare i8 @llvm.cttz.i8 (i8 <src>)
5786 declare i16 @llvm.cttz.i16(i16 <src>)
5787 declare i32 @llvm.cttz.i32(i32 <src>)
5788 declare i64 @llvm.cttz.i64(i64 <src>)
5789 declare i256 @llvm.cttz.i256(i256 <src>)
5795 The '<tt>llvm.cttz</tt>' family of intrinsic functions counts the number of
5802 The only argument is the value to be counted. The argument may be of any
5803 integer type. The return type must match the argument type.
5809 The '<tt>llvm.cttz</tt>' intrinsic counts the trailing (least significant) zeros
5810 in a variable. If the src == 0 then the result is the size in bits of the type
5811 of src. For example, <tt>llvm.cttz(2) = 1</tt>.
5815 <!-- _______________________________________________________________________ -->
5816 <div class="doc_subsubsection">
5817 <a name="int_part_select">'<tt>llvm.part.select.*</tt>' Intrinsic</a>
5820 <div class="doc_text">
5823 <p>This is an overloaded intrinsic. You can use <tt>llvm.part.select</tt>
5824 on any integer bit width.</p>
5826 declare i17 @llvm.part.select.i17 (i17 %val, i32 %loBit, i32 %hiBit)
5827 declare i29 @llvm.part.select.i29 (i29 %val, i32 %loBit, i32 %hiBit)
5831 <p>The '<tt>llvm.part.select</tt>' family of intrinsic functions selects a
5832 range of bits from an integer value and returns them in the same bit width as
5833 the original value.</p>
5836 <p>The first argument, <tt>%val</tt> and the result may be integer types of
5837 any bit width but they must have the same bit width. The second and third
5838 arguments must be <tt>i32</tt> type since they specify only a bit index.</p>
5841 <p>The operation of the '<tt>llvm.part.select</tt>' intrinsic has two modes
5842 of operation: forwards and reverse. If <tt>%loBit</tt> is greater than
5843 <tt>%hiBits</tt> then the intrinsic operates in reverse mode. Otherwise it
5844 operates in forward mode.</p>
5845 <p>In forward mode, this intrinsic is the equivalent of shifting <tt>%val</tt>
5846 right by <tt>%loBit</tt> bits and then ANDing it with a mask with
5847 only the <tt>%hiBit - %loBit</tt> bits set, as follows:</p>
5849 <li>The <tt>%val</tt> is shifted right (LSHR) by the number of bits specified
5850 by <tt>%loBits</tt>. This normalizes the value to the low order bits.</li>
5851 <li>The <tt>%loBits</tt> value is subtracted from the <tt>%hiBits</tt> value
5852 to determine the number of bits to retain.</li>
5853 <li>A mask of the retained bits is created by shifting a -1 value.</li>
5854 <li>The mask is ANDed with <tt>%val</tt> to produce the result.</li>
5856 <p>In reverse mode, a similar computation is made except that the bits are
5857 returned in the reverse order. So, for example, if <tt>X</tt> has the value
5858 <tt>i16 0x0ACF (101011001111)</tt> and we apply
5859 <tt>part.select(i16 X, 8, 3)</tt> to it, we get back the value
5860 <tt>i16 0x0026 (000000100110)</tt>.</p>
5863 <div class="doc_subsubsection">
5864 <a name="int_part_set">'<tt>llvm.part.set.*</tt>' Intrinsic</a>
5867 <div class="doc_text">
5870 <p>This is an overloaded intrinsic. You can use <tt>llvm.part.set</tt>
5871 on any integer bit width.</p>
5873 declare i17 @llvm.part.set.i17.i9 (i17 %val, i9 %repl, i32 %lo, i32 %hi)
5874 declare i29 @llvm.part.set.i29.i9 (i29 %val, i9 %repl, i32 %lo, i32 %hi)
5878 <p>The '<tt>llvm.part.set</tt>' family of intrinsic functions replaces a range
5879 of bits in an integer value with another integer value. It returns the integer
5880 with the replaced bits.</p>
5883 <p>The first argument, <tt>%val</tt>, and the result may be integer types of
5884 any bit width, but they must have the same bit width. <tt>%val</tt> is the value
5885 whose bits will be replaced. The second argument, <tt>%repl</tt> may be an
5886 integer of any bit width. The third and fourth arguments must be <tt>i32</tt>
5887 type since they specify only a bit index.</p>
5890 <p>The operation of the '<tt>llvm.part.set</tt>' intrinsic has two modes
5891 of operation: forwards and reverse. If <tt>%lo</tt> is greater than
5892 <tt>%hi</tt> then the intrinsic operates in reverse mode. Otherwise it
5893 operates in forward mode.</p>
5895 <p>For both modes, the <tt>%repl</tt> value is prepared for use by either
5896 truncating it down to the size of the replacement area or zero extending it
5897 up to that size.</p>
5899 <p>In forward mode, the bits between <tt>%lo</tt> and <tt>%hi</tt> (inclusive)
5900 are replaced with corresponding bits from <tt>%repl</tt>. That is the 0th bit
5901 in <tt>%repl</tt> replaces the <tt>%lo</tt>th bit in <tt>%val</tt> and etc. up
5902 to the <tt>%hi</tt>th bit.</p>
5904 <p>In reverse mode, a similar computation is made except that the bits are
5905 reversed. That is, the <tt>0</tt>th bit in <tt>%repl</tt> replaces the
5906 <tt>%hi</tt> bit in <tt>%val</tt> and etc. down to the <tt>%lo</tt>th bit.</p>
5911 llvm.part.set(0xFFFF, 0, 4, 7) -> 0xFF0F
5912 llvm.part.set(0xFFFF, 0, 7, 4) -> 0xFF0F
5913 llvm.part.set(0xFFFF, 1, 7, 4) -> 0xFF8F
5914 llvm.part.set(0xFFFF, F, 8, 3) -> 0xFFE7
5915 llvm.part.set(0xFFFF, 0, 3, 8) -> 0xFE07
5920 <!-- ======================================================================= -->
5921 <div class="doc_subsection">
5922 <a name="int_overflow">Arithmetic with Overflow Intrinsics</a>
5925 <div class="doc_text">
5927 LLVM provides intrinsics for some arithmetic with overflow operations.
5932 <!-- _______________________________________________________________________ -->
5933 <div class="doc_subsubsection">
5934 <a name="int_sadd_overflow">'<tt>llvm.sadd.with.overflow.*</tt>' Intrinsics</a>
5937 <div class="doc_text">
5941 <p>This is an overloaded intrinsic. You can use <tt>llvm.sadd.with.overflow</tt>
5942 on any integer bit width.</p>
5945 declare {i16, i1} @llvm.sadd.with.overflow.i16(i16 %a, i16 %b)
5946 declare {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
5947 declare {i64, i1} @llvm.sadd.with.overflow.i64(i64 %a, i64 %b)
5952 <p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
5953 a signed addition of the two arguments, and indicate whether an overflow
5954 occurred during the signed summation.</p>
5958 <p>The arguments (%a and %b) and the first element of the result structure may
5959 be of integer types of any bit width, but they must have the same bit width. The
5960 second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
5961 and <tt>%b</tt> are the two values that will undergo signed addition.</p>
5965 <p>The '<tt>llvm.sadd.with.overflow</tt>' family of intrinsic functions perform
5966 a signed addition of the two variables. They return a structure — the
5967 first element of which is the signed summation, and the second element of which
5968 is a bit specifying if the signed summation resulted in an overflow.</p>
5972 %res = call {i32, i1} @llvm.sadd.with.overflow.i32(i32 %a, i32 %b)
5973 %sum = extractvalue {i32, i1} %res, 0
5974 %obit = extractvalue {i32, i1} %res, 1
5975 br i1 %obit, label %overflow, label %normal
5980 <!-- _______________________________________________________________________ -->
5981 <div class="doc_subsubsection">
5982 <a name="int_uadd_overflow">'<tt>llvm.uadd.with.overflow.*</tt>' Intrinsics</a>
5985 <div class="doc_text">
5989 <p>This is an overloaded intrinsic. You can use <tt>llvm.uadd.with.overflow</tt>
5990 on any integer bit width.</p>
5993 declare {i16, i1} @llvm.uadd.with.overflow.i16(i16 %a, i16 %b)
5994 declare {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
5995 declare {i64, i1} @llvm.uadd.with.overflow.i64(i64 %a, i64 %b)
6000 <p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
6001 an unsigned addition of the two arguments, and indicate whether a carry occurred
6002 during the unsigned summation.</p>
6006 <p>The arguments (%a and %b) and the first element of the result structure may
6007 be of integer types of any bit width, but they must have the same bit width. The
6008 second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
6009 and <tt>%b</tt> are the two values that will undergo unsigned addition.</p>
6013 <p>The '<tt>llvm.uadd.with.overflow</tt>' family of intrinsic functions perform
6014 an unsigned addition of the two arguments. They return a structure — the
6015 first element of which is the sum, and the second element of which is a bit
6016 specifying if the unsigned summation resulted in a carry.</p>
6020 %res = call {i32, i1} @llvm.uadd.with.overflow.i32(i32 %a, i32 %b)
6021 %sum = extractvalue {i32, i1} %res, 0
6022 %obit = extractvalue {i32, i1} %res, 1
6023 br i1 %obit, label %carry, label %normal
6028 <!-- _______________________________________________________________________ -->
6029 <div class="doc_subsubsection">
6030 <a name="int_ssub_overflow">'<tt>llvm.ssub.with.overflow.*</tt>' Intrinsics</a>
6033 <div class="doc_text">
6037 <p>This is an overloaded intrinsic. You can use <tt>llvm.ssub.with.overflow</tt>
6038 on any integer bit width.</p>
6041 declare {i16, i1} @llvm.ssub.with.overflow.i16(i16 %a, i16 %b)
6042 declare {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6043 declare {i64, i1} @llvm.ssub.with.overflow.i64(i64 %a, i64 %b)
6048 <p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
6049 a signed subtraction of the two arguments, and indicate whether an overflow
6050 occurred during the signed subtraction.</p>
6054 <p>The arguments (%a and %b) and the first element of the result structure may
6055 be of integer types of any bit width, but they must have the same bit width. The
6056 second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
6057 and <tt>%b</tt> are the two values that will undergo signed subtraction.</p>
6061 <p>The '<tt>llvm.ssub.with.overflow</tt>' family of intrinsic functions perform
6062 a signed subtraction of the two arguments. They return a structure — the
6063 first element of which is the subtraction, and the second element of which is a bit
6064 specifying if the signed subtraction resulted in an overflow.</p>
6068 %res = call {i32, i1} @llvm.ssub.with.overflow.i32(i32 %a, i32 %b)
6069 %sum = extractvalue {i32, i1} %res, 0
6070 %obit = extractvalue {i32, i1} %res, 1
6071 br i1 %obit, label %overflow, label %normal
6076 <!-- _______________________________________________________________________ -->
6077 <div class="doc_subsubsection">
6078 <a name="int_usub_overflow">'<tt>llvm.usub.with.overflow.*</tt>' Intrinsics</a>
6081 <div class="doc_text">
6085 <p>This is an overloaded intrinsic. You can use <tt>llvm.usub.with.overflow</tt>
6086 on any integer bit width.</p>
6089 declare {i16, i1} @llvm.usub.with.overflow.i16(i16 %a, i16 %b)
6090 declare {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6091 declare {i64, i1} @llvm.usub.with.overflow.i64(i64 %a, i64 %b)
6096 <p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
6097 an unsigned subtraction of the two arguments, and indicate whether an overflow
6098 occurred during the unsigned subtraction.</p>
6102 <p>The arguments (%a and %b) and the first element of the result structure may
6103 be of integer types of any bit width, but they must have the same bit width. The
6104 second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
6105 and <tt>%b</tt> are the two values that will undergo unsigned subtraction.</p>
6109 <p>The '<tt>llvm.usub.with.overflow</tt>' family of intrinsic functions perform
6110 an unsigned subtraction of the two arguments. They return a structure — the
6111 first element of which is the subtraction, and the second element of which is a bit
6112 specifying if the unsigned subtraction resulted in an overflow.</p>
6116 %res = call {i32, i1} @llvm.usub.with.overflow.i32(i32 %a, i32 %b)
6117 %sum = extractvalue {i32, i1} %res, 0
6118 %obit = extractvalue {i32, i1} %res, 1
6119 br i1 %obit, label %overflow, label %normal
6124 <!-- _______________________________________________________________________ -->
6125 <div class="doc_subsubsection">
6126 <a name="int_smul_overflow">'<tt>llvm.smul.with.overflow.*</tt>' Intrinsics</a>
6129 <div class="doc_text">
6133 <p>This is an overloaded intrinsic. You can use <tt>llvm.smul.with.overflow</tt>
6134 on any integer bit width.</p>
6137 declare {i16, i1} @llvm.smul.with.overflow.i16(i16 %a, i16 %b)
6138 declare {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6139 declare {i64, i1} @llvm.smul.with.overflow.i64(i64 %a, i64 %b)
6144 <p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
6145 a signed multiplication of the two arguments, and indicate whether an overflow
6146 occurred during the signed multiplication.</p>
6150 <p>The arguments (%a and %b) and the first element of the result structure may
6151 be of integer types of any bit width, but they must have the same bit width. The
6152 second element of the result structure must be of type <tt>i1</tt>. <tt>%a</tt>
6153 and <tt>%b</tt> are the two values that will undergo signed multiplication.</p>
6157 <p>The '<tt>llvm.smul.with.overflow</tt>' family of intrinsic functions perform
6158 a signed multiplication of the two arguments. They return a structure —
6159 the first element of which is the multiplication, and the second element of
6160 which is a bit specifying if the signed multiplication resulted in an
6165 %res = call {i32, i1} @llvm.smul.with.overflow.i32(i32 %a, i32 %b)
6166 %sum = extractvalue {i32, i1} %res, 0
6167 %obit = extractvalue {i32, i1} %res, 1
6168 br i1 %obit, label %overflow, label %normal
6173 <!-- ======================================================================= -->
6174 <div class="doc_subsection">
6175 <a name="int_debugger">Debugger Intrinsics</a>
6178 <div class="doc_text">
6180 The LLVM debugger intrinsics (which all start with <tt>llvm.dbg.</tt> prefix),
6181 are described in the <a
6182 href="SourceLevelDebugging.html#format_common_intrinsics">LLVM Source Level
6183 Debugging</a> document.
6188 <!-- ======================================================================= -->
6189 <div class="doc_subsection">
6190 <a name="int_eh">Exception Handling Intrinsics</a>
6193 <div class="doc_text">
6194 <p> The LLVM exception handling intrinsics (which all start with
6195 <tt>llvm.eh.</tt> prefix), are described in the <a
6196 href="ExceptionHandling.html#format_common_intrinsics">LLVM Exception
6197 Handling</a> document. </p>
6200 <!-- ======================================================================= -->
6201 <div class="doc_subsection">
6202 <a name="int_trampoline">Trampoline Intrinsic</a>
6205 <div class="doc_text">
6207 This intrinsic makes it possible to excise one parameter, marked with
6208 the <tt>nest</tt> attribute, from a function. The result is a callable
6209 function pointer lacking the nest parameter - the caller does not need
6210 to provide a value for it. Instead, the value to use is stored in
6211 advance in a "trampoline", a block of memory usually allocated
6212 on the stack, which also contains code to splice the nest value into the
6213 argument list. This is used to implement the GCC nested function address
6217 For example, if the function is
6218 <tt>i32 f(i8* nest %c, i32 %x, i32 %y)</tt> then the resulting function
6219 pointer has signature <tt>i32 (i32, i32)*</tt>. It can be created as follows:</p>
6221 %tramp = alloca [10 x i8], align 4 ; size and alignment only correct for X86
6222 %tramp1 = getelementptr [10 x i8]* %tramp, i32 0, i32 0
6223 %p = call i8* @llvm.init.trampoline( i8* %tramp1, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
6224 %fp = bitcast i8* %p to i32 (i32, i32)*
6226 <p>The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent
6227 to <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.</p>
6230 <!-- _______________________________________________________________________ -->
6231 <div class="doc_subsubsection">
6232 <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
6234 <div class="doc_text">
6237 declare i8* @llvm.init.trampoline(i8* <tramp>, i8* <func>, i8* <nval>)
6241 This fills the memory pointed to by <tt>tramp</tt> with code
6242 and returns a function pointer suitable for executing it.
6246 The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
6247 pointers. The <tt>tramp</tt> argument must point to a sufficiently large
6248 and sufficiently aligned block of memory; this memory is written to by the
6249 intrinsic. Note that the size and the alignment are target-specific - LLVM
6250 currently provides no portable way of determining them, so a front-end that
6251 generates this intrinsic needs to have some target-specific knowledge.
6252 The <tt>func</tt> argument must hold a function bitcast to an <tt>i8*</tt>.
6256 The block of memory pointed to by <tt>tramp</tt> is filled with target
6257 dependent code, turning it into a function. A pointer to this function is
6258 returned, but needs to be bitcast to an
6259 <a href="#int_trampoline">appropriate function pointer type</a>
6260 before being called. The new function's signature is the same as that of
6261 <tt>func</tt> with any arguments marked with the <tt>nest</tt> attribute
6262 removed. At most one such <tt>nest</tt> argument is allowed, and it must be
6263 of pointer type. Calling the new function is equivalent to calling
6264 <tt>func</tt> with the same argument list, but with <tt>nval</tt> used for the
6265 missing <tt>nest</tt> argument. If, after calling
6266 <tt>llvm.init.trampoline</tt>, the memory pointed to by <tt>tramp</tt> is
6267 modified, then the effect of any later call to the returned function pointer is
6272 <!-- ======================================================================= -->
6273 <div class="doc_subsection">
6274 <a name="int_atomics">Atomic Operations and Synchronization Intrinsics</a>
6277 <div class="doc_text">
6279 These intrinsic functions expand the "universal IR" of LLVM to represent
6280 hardware constructs for atomic operations and memory synchronization. This
6281 provides an interface to the hardware, not an interface to the programmer. It
6282 is aimed at a low enough level to allow any programming models or APIs
6283 (Application Programming Interfaces) which
6284 need atomic behaviors to map cleanly onto it. It is also modeled primarily on
6285 hardware behavior. Just as hardware provides a "universal IR" for source
6286 languages, it also provides a starting point for developing a "universal"
6287 atomic operation and synchronization IR.
6290 These do <em>not</em> form an API such as high-level threading libraries,
6291 software transaction memory systems, atomic primitives, and intrinsic
6292 functions as found in BSD, GNU libc, atomic_ops, APR, and other system and
6293 application libraries. The hardware interface provided by LLVM should allow
6294 a clean implementation of all of these APIs and parallel programming models.
6295 No one model or paradigm should be selected above others unless the hardware
6296 itself ubiquitously does so.
6301 <!-- _______________________________________________________________________ -->
6302 <div class="doc_subsubsection">
6303 <a name="int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a>
6305 <div class="doc_text">
6308 declare void @llvm.memory.barrier( i1 <ll>, i1 <ls>, i1 <sl>, i1 <ss>,
6314 The <tt>llvm.memory.barrier</tt> intrinsic guarantees ordering between
6315 specific pairs of memory access types.
6319 The <tt>llvm.memory.barrier</tt> intrinsic requires five boolean arguments.
6320 The first four arguments enables a specific barrier as listed below. The fith
6321 argument specifies that the barrier applies to io or device or uncached memory.
6325 <li><tt>ll</tt>: load-load barrier</li>
6326 <li><tt>ls</tt>: load-store barrier</li>
6327 <li><tt>sl</tt>: store-load barrier</li>
6328 <li><tt>ss</tt>: store-store barrier</li>
6329 <li><tt>device</tt>: barrier applies to device and uncached memory also.</li>
6333 This intrinsic causes the system to enforce some ordering constraints upon
6334 the loads and stores of the program. This barrier does not indicate
6335 <em>when</em> any events will occur, it only enforces an <em>order</em> in
6336 which they occur. For any of the specified pairs of load and store operations
6337 (f.ex. load-load, or store-load), all of the first operations preceding the
6338 barrier will complete before any of the second operations succeeding the
6339 barrier begin. Specifically the semantics for each pairing is as follows:
6342 <li><tt>ll</tt>: All loads before the barrier must complete before any load
6343 after the barrier begins.</li>
6345 <li><tt>ls</tt>: All loads before the barrier must complete before any
6346 store after the barrier begins.</li>
6347 <li><tt>ss</tt>: All stores before the barrier must complete before any
6348 store after the barrier begins.</li>
6349 <li><tt>sl</tt>: All stores before the barrier must complete before any
6350 load after the barrier begins.</li>
6353 These semantics are applied with a logical "and" behavior when more than one
6354 is enabled in a single memory barrier intrinsic.
6357 Backends may implement stronger barriers than those requested when they do not
6358 support as fine grained a barrier as requested. Some architectures do not
6359 need all types of barriers and on such architectures, these become noops.
6366 %result1 = load i32* %ptr <i>; yields {i32}:result1 = 4</i>
6367 call void @llvm.memory.barrier( i1 false, i1 true, i1 false, i1 false )
6368 <i>; guarantee the above finishes</i>
6369 store i32 8, %ptr <i>; before this begins</i>
6373 <!-- _______________________________________________________________________ -->
6374 <div class="doc_subsubsection">
6375 <a name="int_atomic_cmp_swap">'<tt>llvm.atomic.cmp.swap.*</tt>' Intrinsic</a>
6377 <div class="doc_text">
6380 This is an overloaded intrinsic. You can use <tt>llvm.atomic.cmp.swap</tt> on
6381 any integer bit width and for different address spaces. Not all targets
6382 support all bit widths however.</p>
6385 declare i8 @llvm.atomic.cmp.swap.i8.p0i8( i8* <ptr>, i8 <cmp>, i8 <val> )
6386 declare i16 @llvm.atomic.cmp.swap.i16.p0i16( i16* <ptr>, i16 <cmp>, i16 <val> )
6387 declare i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* <ptr>, i32 <cmp>, i32 <val> )
6388 declare i64 @llvm.atomic.cmp.swap.i64.p0i64( i64* <ptr>, i64 <cmp>, i64 <val> )
6393 This loads a value in memory and compares it to a given value. If they are
6394 equal, it stores a new value into the memory.
6398 The <tt>llvm.atomic.cmp.swap</tt> intrinsic takes three arguments. The result as
6399 well as both <tt>cmp</tt> and <tt>val</tt> must be integer values with the
6400 same bit width. The <tt>ptr</tt> argument must be a pointer to a value of
6401 this integer type. While any bit width integer may be used, targets may only
6402 lower representations they support in hardware.
6407 This entire intrinsic must be executed atomically. It first loads the value
6408 in memory pointed to by <tt>ptr</tt> and compares it with the value
6409 <tt>cmp</tt>. If they are equal, <tt>val</tt> is stored into the memory. The
6410 loaded value is yielded in all cases. This provides the equivalent of an
6411 atomic compare-and-swap operation within the SSA framework.
6419 %val1 = add i32 4, 4
6420 %result1 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 4, %val1 )
6421 <i>; yields {i32}:result1 = 4</i>
6422 %stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6423 %memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6425 %val2 = add i32 1, 1
6426 %result2 = call i32 @llvm.atomic.cmp.swap.i32.p0i32( i32* %ptr, i32 5, %val2 )
6427 <i>; yields {i32}:result2 = 8</i>
6428 %stored2 = icmp eq i32 %result2, 5 <i>; yields {i1}:stored2 = false</i>
6430 %memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 8</i>
6434 <!-- _______________________________________________________________________ -->
6435 <div class="doc_subsubsection">
6436 <a name="int_atomic_swap">'<tt>llvm.atomic.swap.*</tt>' Intrinsic</a>
6438 <div class="doc_text">
6442 This is an overloaded intrinsic. You can use <tt>llvm.atomic.swap</tt> on any
6443 integer bit width. Not all targets support all bit widths however.</p>
6445 declare i8 @llvm.atomic.swap.i8.p0i8( i8* <ptr>, i8 <val> )
6446 declare i16 @llvm.atomic.swap.i16.p0i16( i16* <ptr>, i16 <val> )
6447 declare i32 @llvm.atomic.swap.i32.p0i32( i32* <ptr>, i32 <val> )
6448 declare i64 @llvm.atomic.swap.i64.p0i64( i64* <ptr>, i64 <val> )
6453 This intrinsic loads the value stored in memory at <tt>ptr</tt> and yields
6454 the value from memory. It then stores the value in <tt>val</tt> in the memory
6460 The <tt>llvm.atomic.swap</tt> intrinsic takes two arguments. Both the
6461 <tt>val</tt> argument and the result must be integers of the same bit width.
6462 The first argument, <tt>ptr</tt>, must be a pointer to a value of this
6463 integer type. The targets may only lower integer representations they
6468 This intrinsic loads the value pointed to by <tt>ptr</tt>, yields it, and
6469 stores <tt>val</tt> back into <tt>ptr</tt> atomically. This provides the
6470 equivalent of an atomic swap operation within the SSA framework.
6478 %val1 = add i32 4, 4
6479 %result1 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val1 )
6480 <i>; yields {i32}:result1 = 4</i>
6481 %stored1 = icmp eq i32 %result1, 4 <i>; yields {i1}:stored1 = true</i>
6482 %memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 8</i>
6484 %val2 = add i32 1, 1
6485 %result2 = call i32 @llvm.atomic.swap.i32.p0i32( i32* %ptr, i32 %val2 )
6486 <i>; yields {i32}:result2 = 8</i>
6488 %stored2 = icmp eq i32 %result2, 8 <i>; yields {i1}:stored2 = true</i>
6489 %memval2 = load i32* %ptr <i>; yields {i32}:memval2 = 2</i>
6493 <!-- _______________________________________________________________________ -->
6494 <div class="doc_subsubsection">
6495 <a name="int_atomic_load_add">'<tt>llvm.atomic.load.add.*</tt>' Intrinsic</a>
6498 <div class="doc_text">
6501 This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.add</tt> on any
6502 integer bit width. Not all targets support all bit widths however.</p>
6504 declare i8 @llvm.atomic.load.add.i8..p0i8( i8* <ptr>, i8 <delta> )
6505 declare i16 @llvm.atomic.load.add.i16..p0i16( i16* <ptr>, i16 <delta> )
6506 declare i32 @llvm.atomic.load.add.i32..p0i32( i32* <ptr>, i32 <delta> )
6507 declare i64 @llvm.atomic.load.add.i64..p0i64( i64* <ptr>, i64 <delta> )
6512 This intrinsic adds <tt>delta</tt> to the value stored in memory at
6513 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.
6518 The intrinsic takes two arguments, the first a pointer to an integer value
6519 and the second an integer value. The result is also an integer value. These
6520 integer types can have any bit width, but they must all have the same bit
6521 width. The targets may only lower integer representations they support.
6525 This intrinsic does a series of operations atomically. It first loads the
6526 value stored at <tt>ptr</tt>. It then adds <tt>delta</tt>, stores the result
6527 to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.
6534 %result1 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 4 )
6535 <i>; yields {i32}:result1 = 4</i>
6536 %result2 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 2 )
6537 <i>; yields {i32}:result2 = 8</i>
6538 %result3 = call i32 @llvm.atomic.load.add.i32.p0i32( i32* %ptr, i32 5 )
6539 <i>; yields {i32}:result3 = 10</i>
6540 %memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 15</i>
6544 <!-- _______________________________________________________________________ -->
6545 <div class="doc_subsubsection">
6546 <a name="int_atomic_load_sub">'<tt>llvm.atomic.load.sub.*</tt>' Intrinsic</a>
6549 <div class="doc_text">
6552 This is an overloaded intrinsic. You can use <tt>llvm.atomic.load.sub</tt> on
6553 any integer bit width and for different address spaces. Not all targets
6554 support all bit widths however.</p>
6556 declare i8 @llvm.atomic.load.sub.i8.p0i32( i8* <ptr>, i8 <delta> )
6557 declare i16 @llvm.atomic.load.sub.i16.p0i32( i16* <ptr>, i16 <delta> )
6558 declare i32 @llvm.atomic.load.sub.i32.p0i32( i32* <ptr>, i32 <delta> )
6559 declare i64 @llvm.atomic.load.sub.i64.p0i32( i64* <ptr>, i64 <delta> )
6564 This intrinsic subtracts <tt>delta</tt> to the value stored in memory at
6565 <tt>ptr</tt>. It yields the original value at <tt>ptr</tt>.
6570 The intrinsic takes two arguments, the first a pointer to an integer value
6571 and the second an integer value. The result is also an integer value. These
6572 integer types can have any bit width, but they must all have the same bit
6573 width. The targets may only lower integer representations they support.
6577 This intrinsic does a series of operations atomically. It first loads the
6578 value stored at <tt>ptr</tt>. It then subtracts <tt>delta</tt>, stores the
6579 result to <tt>ptr</tt>. It yields the original value stored at <tt>ptr</tt>.
6586 %result1 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 4 )
6587 <i>; yields {i32}:result1 = 8</i>
6588 %result2 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 2 )
6589 <i>; yields {i32}:result2 = 4</i>
6590 %result3 = call i32 @llvm.atomic.load.sub.i32.p0i32( i32* %ptr, i32 5 )
6591 <i>; yields {i32}:result3 = 2</i>
6592 %memval1 = load i32* %ptr <i>; yields {i32}:memval1 = -3</i>
6596 <!-- _______________________________________________________________________ -->
6597 <div class="doc_subsubsection">
6598 <a name="int_atomic_load_and">'<tt>llvm.atomic.load.and.*</tt>' Intrinsic</a><br>
6599 <a name="int_atomic_load_nand">'<tt>llvm.atomic.load.nand.*</tt>' Intrinsic</a><br>
6600 <a name="int_atomic_load_or">'<tt>llvm.atomic.load.or.*</tt>' Intrinsic</a><br>
6601 <a name="int_atomic_load_xor">'<tt>llvm.atomic.load.xor.*</tt>' Intrinsic</a><br>
6604 <div class="doc_text">
6607 These are overloaded intrinsics. You can use <tt>llvm.atomic.load_and</tt>,
6608 <tt>llvm.atomic.load_nand</tt>, <tt>llvm.atomic.load_or</tt>, and
6609 <tt>llvm.atomic.load_xor</tt> on any integer bit width and for different
6610 address spaces. Not all targets support all bit widths however.</p>
6612 declare i8 @llvm.atomic.load.and.i8.p0i8( i8* <ptr>, i8 <delta> )
6613 declare i16 @llvm.atomic.load.and.i16.p0i16( i16* <ptr>, i16 <delta> )
6614 declare i32 @llvm.atomic.load.and.i32.p0i32( i32* <ptr>, i32 <delta> )
6615 declare i64 @llvm.atomic.load.and.i64.p0i64( i64* <ptr>, i64 <delta> )
6620 declare i8 @llvm.atomic.load.or.i8.p0i8( i8* <ptr>, i8 <delta> )
6621 declare i16 @llvm.atomic.load.or.i16.p0i16( i16* <ptr>, i16 <delta> )
6622 declare i32 @llvm.atomic.load.or.i32.p0i32( i32* <ptr>, i32 <delta> )
6623 declare i64 @llvm.atomic.load.or.i64.p0i64( i64* <ptr>, i64 <delta> )
6628 declare i8 @llvm.atomic.load.nand.i8.p0i32( i8* <ptr>, i8 <delta> )
6629 declare i16 @llvm.atomic.load.nand.i16.p0i32( i16* <ptr>, i16 <delta> )
6630 declare i32 @llvm.atomic.load.nand.i32.p0i32( i32* <ptr>, i32 <delta> )
6631 declare i64 @llvm.atomic.load.nand.i64.p0i32( i64* <ptr>, i64 <delta> )
6636 declare i8 @llvm.atomic.load.xor.i8.p0i32( i8* <ptr>, i8 <delta> )
6637 declare i16 @llvm.atomic.load.xor.i16.p0i32( i16* <ptr>, i16 <delta> )
6638 declare i32 @llvm.atomic.load.xor.i32.p0i32( i32* <ptr>, i32 <delta> )
6639 declare i64 @llvm.atomic.load.xor.i64.p0i32( i64* <ptr>, i64 <delta> )
6644 These intrinsics bitwise the operation (and, nand, or, xor) <tt>delta</tt> to
6645 the value stored in memory at <tt>ptr</tt>. It yields the original value
6651 These intrinsics take two arguments, the first a pointer to an integer value
6652 and the second an integer value. The result is also an integer value. These
6653 integer types can have any bit width, but they must all have the same bit
6654 width. The targets may only lower integer representations they support.
6658 These intrinsics does a series of operations atomically. They first load the
6659 value stored at <tt>ptr</tt>. They then do the bitwise operation
6660 <tt>delta</tt>, store the result to <tt>ptr</tt>. They yield the original
6661 value stored at <tt>ptr</tt>.
6667 store i32 0x0F0F, %ptr
6668 %result0 = call i32 @llvm.atomic.load.nand.i32.p0i32( i32* %ptr, i32 0xFF )
6669 <i>; yields {i32}:result0 = 0x0F0F</i>
6670 %result1 = call i32 @llvm.atomic.load.and.i32.p0i32( i32* %ptr, i32 0xFF )
6671 <i>; yields {i32}:result1 = 0xFFFFFFF0</i>
6672 %result2 = call i32 @llvm.atomic.load.or.i32.p0i32( i32* %ptr, i32 0F )
6673 <i>; yields {i32}:result2 = 0xF0</i>
6674 %result3 = call i32 @llvm.atomic.load.xor.i32.p0i32( i32* %ptr, i32 0F )
6675 <i>; yields {i32}:result3 = FF</i>
6676 %memval1 = load i32* %ptr <i>; yields {i32}:memval1 = F0</i>
6681 <!-- _______________________________________________________________________ -->
6682 <div class="doc_subsubsection">
6683 <a name="int_atomic_load_max">'<tt>llvm.atomic.load.max.*</tt>' Intrinsic</a><br>
6684 <a name="int_atomic_load_min">'<tt>llvm.atomic.load.min.*</tt>' Intrinsic</a><br>
6685 <a name="int_atomic_load_umax">'<tt>llvm.atomic.load.umax.*</tt>' Intrinsic</a><br>
6686 <a name="int_atomic_load_umin">'<tt>llvm.atomic.load.umin.*</tt>' Intrinsic</a><br>
6689 <div class="doc_text">
6692 These are overloaded intrinsics. You can use <tt>llvm.atomic.load_max</tt>,
6693 <tt>llvm.atomic.load_min</tt>, <tt>llvm.atomic.load_umax</tt>, and
6694 <tt>llvm.atomic.load_umin</tt> on any integer bit width and for different
6695 address spaces. Not all targets
6696 support all bit widths however.</p>
6698 declare i8 @llvm.atomic.load.max.i8.p0i8( i8* <ptr>, i8 <delta> )
6699 declare i16 @llvm.atomic.load.max.i16.p0i16( i16* <ptr>, i16 <delta> )
6700 declare i32 @llvm.atomic.load.max.i32.p0i32( i32* <ptr>, i32 <delta> )
6701 declare i64 @llvm.atomic.load.max.i64.p0i64( i64* <ptr>, i64 <delta> )
6706 declare i8 @llvm.atomic.load.min.i8.p0i8( i8* <ptr>, i8 <delta> )
6707 declare i16 @llvm.atomic.load.min.i16.p0i16( i16* <ptr>, i16 <delta> )
6708 declare i32 @llvm.atomic.load.min.i32..p0i32( i32* <ptr>, i32 <delta> )
6709 declare i64 @llvm.atomic.load.min.i64..p0i64( i64* <ptr>, i64 <delta> )
6714 declare i8 @llvm.atomic.load.umax.i8.p0i8( i8* <ptr>, i8 <delta> )
6715 declare i16 @llvm.atomic.load.umax.i16.p0i16( i16* <ptr>, i16 <delta> )
6716 declare i32 @llvm.atomic.load.umax.i32.p0i32( i32* <ptr>, i32 <delta> )
6717 declare i64 @llvm.atomic.load.umax.i64.p0i64( i64* <ptr>, i64 <delta> )
6722 declare i8 @llvm.atomic.load.umin.i8..p0i8( i8* <ptr>, i8 <delta> )
6723 declare i16 @llvm.atomic.load.umin.i16.p0i16( i16* <ptr>, i16 <delta> )
6724 declare i32 @llvm.atomic.load.umin.i32..p0i32( i32* <ptr>, i32 <delta> )
6725 declare i64 @llvm.atomic.load.umin.i64..p0i64( i64* <ptr>, i64 <delta> )
6730 These intrinsics takes the signed or unsigned minimum or maximum of
6731 <tt>delta</tt> and the value stored in memory at <tt>ptr</tt>. It yields the
6732 original value at <tt>ptr</tt>.
6737 These intrinsics take two arguments, the first a pointer to an integer value
6738 and the second an integer value. The result is also an integer value. These
6739 integer types can have any bit width, but they must all have the same bit
6740 width. The targets may only lower integer representations they support.
6744 These intrinsics does a series of operations atomically. They first load the
6745 value stored at <tt>ptr</tt>. They then do the signed or unsigned min or max
6746 <tt>delta</tt> and the value, store the result to <tt>ptr</tt>. They yield
6747 the original value stored at <tt>ptr</tt>.
6754 %result0 = call i32 @llvm.atomic.load.min.i32.p0i32( i32* %ptr, i32 -2 )
6755 <i>; yields {i32}:result0 = 7</i>
6756 %result1 = call i32 @llvm.atomic.load.max.i32.p0i32( i32* %ptr, i32 8 )
6757 <i>; yields {i32}:result1 = -2</i>
6758 %result2 = call i32 @llvm.atomic.load.umin.i32.p0i32( i32* %ptr, i32 10 )
6759 <i>; yields {i32}:result2 = 8</i>
6760 %result3 = call i32 @llvm.atomic.load.umax.i32.p0i32( i32* %ptr, i32 30 )
6761 <i>; yields {i32}:result3 = 8</i>
6762 %memval1 = load i32* %ptr <i>; yields {i32}:memval1 = 30</i>
6766 <!-- ======================================================================= -->
6767 <div class="doc_subsection">
6768 <a name="int_general">General Intrinsics</a>
6771 <div class="doc_text">
6772 <p> This class of intrinsics is designed to be generic and has
6773 no specific purpose. </p>
6776 <!-- _______________________________________________________________________ -->
6777 <div class="doc_subsubsection">
6778 <a name="int_var_annotation">'<tt>llvm.var.annotation</tt>' Intrinsic</a>
6781 <div class="doc_text">
6785 declare void @llvm.var.annotation(i8* <val>, i8* <str>, i8* <str>, i32 <int> )
6791 The '<tt>llvm.var.annotation</tt>' intrinsic
6797 The first argument is a pointer to a value, the second is a pointer to a
6798 global string, the third is a pointer to a global string which is the source
6799 file name, and the last argument is the line number.
6805 This intrinsic allows annotation of local variables with arbitrary strings.
6806 This can be useful for special purpose optimizations that want to look for these
6807 annotations. These have no other defined use, they are ignored by code
6808 generation and optimization.
6812 <!-- _______________________________________________________________________ -->
6813 <div class="doc_subsubsection">
6814 <a name="int_annotation">'<tt>llvm.annotation.*</tt>' Intrinsic</a>
6817 <div class="doc_text">
6820 <p>This is an overloaded intrinsic. You can use '<tt>llvm.annotation</tt>' on
6821 any integer bit width.
6824 declare i8 @llvm.annotation.i8(i8 <val>, i8* <str>, i8* <str>, i32 <int> )
6825 declare i16 @llvm.annotation.i16(i16 <val>, i8* <str>, i8* <str>, i32 <int> )
6826 declare i32 @llvm.annotation.i32(i32 <val>, i8* <str>, i8* <str>, i32 <int> )
6827 declare i64 @llvm.annotation.i64(i64 <val>, i8* <str>, i8* <str>, i32 <int> )
6828 declare i256 @llvm.annotation.i256(i256 <val>, i8* <str>, i8* <str>, i32 <int> )
6834 The '<tt>llvm.annotation</tt>' intrinsic.
6840 The first argument is an integer value (result of some expression),
6841 the second is a pointer to a global string, the third is a pointer to a global
6842 string which is the source file name, and the last argument is the line number.
6843 It returns the value of the first argument.
6849 This intrinsic allows annotations to be put on arbitrary expressions
6850 with arbitrary strings. This can be useful for special purpose optimizations
6851 that want to look for these annotations. These have no other defined use, they
6852 are ignored by code generation and optimization.
6856 <!-- _______________________________________________________________________ -->
6857 <div class="doc_subsubsection">
6858 <a name="int_trap">'<tt>llvm.trap</tt>' Intrinsic</a>
6861 <div class="doc_text">
6865 declare void @llvm.trap()
6871 The '<tt>llvm.trap</tt>' intrinsic
6883 This intrinsics is lowered to the target dependent trap instruction. If the
6884 target does not have a trap instruction, this intrinsic will be lowered to the
6885 call of the abort() function.
6889 <!-- _______________________________________________________________________ -->
6890 <div class="doc_subsubsection">
6891 <a name="int_stackprotector">'<tt>llvm.stackprotector</tt>' Intrinsic</a>
6893 <div class="doc_text">
6896 declare void @llvm.stackprotector( i8* <guard>, i8** <slot> )
6901 The <tt>llvm.stackprotector</tt> intrinsic takes the <tt>guard</tt> and stores
6902 it onto the stack at <tt>slot</tt>. The stack slot is adjusted to ensure that
6903 it is placed on the stack before local variables.
6907 The <tt>llvm.stackprotector</tt> intrinsic requires two pointer arguments. The
6908 first argument is the value loaded from the stack guard
6909 <tt>@__stack_chk_guard</tt>. The second variable is an <tt>alloca</tt> that
6910 has enough space to hold the value of the guard.
6914 This intrinsic causes the prologue/epilogue inserter to force the position of
6915 the <tt>AllocaInst</tt> stack slot to be before local variables on the
6916 stack. This is to ensure that if a local variable on the stack is overwritten,
6917 it will destroy the value of the guard. When the function exits, the guard on
6918 the stack is checked against the original guard. If they're different, then
6919 the program aborts by calling the <tt>__stack_chk_fail()</tt> function.
6923 <!-- *********************************************************************** -->
6926 <a href="http://jigsaw.w3.org/css-validator/check/referer"><img
6927 src="http://jigsaw.w3.org/css-validator/images/vcss-blue" alt="Valid CSS"></a>
6928 <a href="http://validator.w3.org/check/referer"><img
6929 src="http://www.w3.org/Icons/valid-html401-blue" alt="Valid HTML 4.01"></a>
6931 <a href="mailto:sabre@nondot.org">Chris Lattner</a><br>
6932 <a href="http://llvm.org">The LLVM Compiler Infrastructure</a><br>
6933 Last modified: $Date$